First Solar, Inc. Aktie

Good afternoon, and welcome to First Solar's Third Quarter 2026 Earnings Conference Call. This call is being webcast live on the Investors section of First Solar's website at investor.firstsolar.com. [Operator Instructions] And please note that today's call is being recorded.

I would now like to turn the call over to your host, Byron Jeffers, Head of Investor Relations.

Good afternoon, and thank you for joining us. With me today are Mark Widmar, Chief Executive Officer; and Alex Bradley, Chief Financial Officer. Mark will begin with an update on the business, followed by Alex with our financial results and outlook.

Before we begin, today's discussion will include forward-looking statements that are subject to risks and uncertainties. Additional information regarding these risks and uncertainties can be found in our earnings press release, other SEC filings and the earnings material available on our Investor Relations website at investor.firstsolar.com. We undertake no obligation to update any of these forward-looking statements as a result of new information or future events.

During today's call, we will also discuss certain non-GAAP financial measures. Reconciliation of these measures to the most directly comparable GAAP measures are provided in our earnings press release and accompanying presentation. We encourage investors to review those reconciliations as non-GAAP measures should not be considered in isolation or a substitute for financial information prepared in accordance with U.S. GAAP.

With that, I'll turn it over to Mark.

Thank you, and good afternoon. Getting on Slide 4. We continue to execute with discipline this past quarter advancing our U.S. manufacturing expansion, care deployment and perovskite R&D program while delivering strong financial results. We generated approximately $1.2 billion in net sales, expanded gross margin to approximately 58% and delivered adjusted EBITDA of $744 million.

Operationally, our continued CuRe deployment across the fleet is being executed through a disciplined learning-based approach that began with our Perrysburg Series 6 conversion in Q1 of this year. That approach now extends to our first Series 7 conversion in our India facility expected to begin in the first quarter of 2027, which marks the start of Series 7 CuRe replication across the fleet.

Beginning in India allows us to first refine the deployment process in a low earnings impact environment before replicating across our U.S. Series 7 fleet. During the subsequent U.S. conversions assuming a supportive trade environment, India front-end output is planned to be flexed to supply U.S. facilities, supporting U.S. back-end throughput and preserving partial domestic content and Section 45x eligibility production during the transition.

Accordingly, India's finished goods production in 2027 is expected to range from 2.2 gigawatts to 2.6 gigawatts. The pace in sequencing subsequent Series 7 U.S. conversions will be informed by learnings from the India launch consistent with the disciplined learning-based approach we have taken across all periods of technology transition.

In South Carolina, our Series 6 finishing facility is being brought online in two phases as part of the integrated manufacturing flow from Vietnam. The first phase started production in Q3. Second phase is expected to reach commercial launch in mid-2027. As we previously shared, we delayed the second phase by approximately one quarter so that Secure could be incorporated at commercial launch. Together, these two phases are expected to result in 2027 South Carolina production of 2.2 to 2.5 gigawatts.

Beyond CuRe, we continue to advance our next-generation technology road map, our Ohio pilot line designed to produce Series 6 size perovskite modules remains on track to reach operational readiness in the first half of 2027. We expect this line will enable field deployment of modules to further capture learnings and validate the manufacturing performance and durability profile of this next-generation thin film technology before we undertake broader scaling decisions.

Our licensing agreement with Cubic PV announced earlier today, together with our previously announced intellectual property agreement with Oxford PV, continues to support the development road map, providing access to foundational perovskite IP as we advance our technology road map.

Turning to the market environment on Slide 5. The Section 232 reclamation issued in August represents a meaningful shift for U.S. solar manufacturing. And we believe First Solar is well positioned to benefit as the shift takes effect. Combined with fundamentals such as continued demand strength from load growth and data center expansions, we see a long-term opportunity to translate this position into earnings growth.

Section 232 represents a structural change in how the United States approaches solar manufacturing. As the U.S. competes for global leadership in artificial intelligence, and the emerging low earth orbit economy, secure and affordable power will be foundational to both. The country cannot afford to depend on adversarial nations for technologies and supply chains increasingly central to our long-term economic prosperity and national security.

As a technology company and the only fully integrated utility scale solar manufacturer without dependency on a Chinese crystal silicon supply chain, we believe we are structurally positioned to benefit from this environment. Note, by 2027, nearly 80% of our global module manufacturing footprint will be located in the U.S.

Long-term demand fundamentals also remained strong. AI-related infrastructure, electrification and the onshoring of U.S. manufacturing are driving significant load growth. These customers need large-scale power quickly and solar plus storage remains one of the fastest and lowest cost solutions. They are increasingly focused on availability, delivery certainty, a bankable warranties and supply chain security attributes we have built our business around.

As it relates to bookings, we believe the market is working through a transition following the Section 232 announcement with significant influence from other key factors. As a reminder, a substantial portion of our existing backlog is comprised of multiyear framework agreements that are intended to supply customers with projects in various stages of development.

While entering into these types of agreements has historically driven significant additions to our backlog, they can also be more challenging for us to plan for and allocate against because of the inherent delivery timing uncertainty associated with projects in early stage of development.

Over the course of this year, while not closing the door on framework agreements with appropriate contractual protections, our customer engagements have shifted more towards project-specific contracting with the 5 projects and delivery schedules. From our perspective, the enhanced certainty surrounding project-specific contracts allows us to optimize module allocation and delivery with a potential trade-off of slowing the pace of an extended multiyear contracted backlog.

At the same time, customers are focused on advancing mid- and late-stage projects where modules have already been secured, while deferring procurement on early-stage projects requiring delivery later in the decade. We see several factors driving this behavior, including increased nimbyism among community residents and growing state and local government opposition.

In addition, the overall inflationary environment, together with the continuously evolving tariff landscape, have challenged near-term project economics and require greater strategic planning on behalf of our development customers to address system cost increases unrelated to module pricing.

As one example, in certain cases, surging battery prices is challenging both projects in construction as well as projects at the PPA bid stage. Lastly, an incomplete regulatory picture with regards to the final FIAT guidance as well as the potential for Section 232 tariffs to continue to evolve, particularly as it relates to minimum import prices and the effectiveness of anti-stocking measures has generated volatility in the module procurement market and inhibited customer clarity as it relates to module pricing discovery.

The impact of these current dynamics on recent bookings is most pronounced from our Malaysia facility. While Vietnam is expected to supply the semi-finished product for our South Carolina finishing facility, Malaysia retains the capacity to produce fully finished international modules. We continue to evaluate demand for this product based on core unit economics, tariffs and trade conditions and customer contracting terms.

We expect the operating decision regarding our Malaysia facility to be binary, resulting in either operations at full or near full capacity of 1.8 gigawatt annually or a shutdown of production with a decision expected in the first half of 2027.

Given Malaysia's relatively small contribution to our manufacturing footprint, limited Series 6 blending opportunity in the U.S. absent a Section 45x credits and freight and tariff burdens, we do not expect either outcome to materially impact our run rate earnings profile.

Against this backdrop, we remain deliberate on bookings, and we are pleased with the results of this approach. Through today in the U.S., our largest market, we have booked approximately 4.9 gigawatts at approximately $0.36 per watt inclusive of ingesters. Moreover, our 44 gigawatt backlog gives us the ability to focus on contracting approaches on two strategic initiatives.

First, we seek to improve pricing on new module sales relative to our current backlog ASP, no pricing will continue to vary by product, Series 6 and Series 7 and by U.S. versus internationally made. Second, we seek to improve contract structures to further mitigate the risk of cost increases beyond our control, strengthening credit support and payment terms and firm up product delivery schedules. Our objective is to optimize across both upfront ASPs as well as long-term earnings profile through disciplined pricing, appropriate contractual protections, strong counterparties and alignment with our manufacturing and delivery strategy.

In summary, our focus on project specific versus framework agreements, customer focus on near-term project execution versus late data procurement, evolving market price discovery and our contracting discipline have moderated recent bookings on a volume basis. But on an economic basis, has generated improved pricing relative to our overall backlog. In addition, we remain focused on our ability to affect our contracting strategic initiatives and are confident in demand for our domestically produced product, particularly for [ 2029 ] and beyond when we have -- available supply.

With that, I turn the call over to Alex to discuss our bookings, financial results and outlook.

Thanks, Mark. Turning to our bookings and backlog on Slide 6. We ended the quarter with 44 gigawatts on to contract, representing approximately $13.5 billion contracted value, exclusive of technology adjusters with scheduled deliveries extending through 2030. Since our last earnings call, we recorded approximately 1.4 gigawatts of additional U.S. gross bookings at an average selling price of approximately $0.36 per watt, inclusive of applicable technology adjusters.

Importantly, substantially all the volume was contracted or in late-stage negotiations and pricing in terms are largely set ahead of the Section 232 announced. In addition, there remains approximately 4 gigawatts of variable volume associated with previously signed contracts, not currently part of the contracted backlog, but pricing reflective of the pre-Section 232 environment, which our customers have the ability to exercise as of the date start. Approximately half of this volume has a decision day within the next 5 months, with the other half spread across the remainder of 2027 and 2028. Should our customers elect to receive this volume, it will be recognized as incremental bookings time of such election.

During the quarter, we also continued customer notifications related to contractual technology adjusters, aligned with our production plan and intended deployment sequence. These notifications represent a meaningful source of incremental backlog value as CuRe deployment advances across the fleet, with the majority of the adjusted value expected to be realized in 2027 and 2028.

In respect to adjust eligible volume, we would note that the ultimate realization of these adjustments is subject to the achievement of the underlying technology milestones, timing of future customer notifications and module delivery schedules and other contractual conditions. And therefore, the final value and timing may differ from the amounts and periods described.

Of that potential adjusted value, approximately 20% is currently associated with 2027 deliveries with the remainder weighted to 2028 and beyond. As we continue to progress customer notifications associated with existing bookings tied to our CuRe deployment sequence, together with new bookings removing the adjusted concepts as we sell our CuRe product with the technology embedded in the base ASP the value of the adjusters will roll off over time.

Turning to Slide 7. Net sales for the third quarter were approximately $1.2 billion, a decrease of approximately 23% year-over-year. The decrease was primarily driven by lower module volumes sold to third parties compared to the prior year period. Gross margin was approximately 58%, up roughly 20 percentage points year-over-year. The year-over-year increase was driven primarily by a higher mix of U.S. produced modules.

Gross margin also benefited from lower sales rates and utilization costs, tariffs and duties, partially offset by lower termination related revenue. Operating expenses were approximately $167 million. Research and development expense was approximately $67 million, reflecting continued investment in our technology road map, including development.

Net income was $571 million, up approximately 25% year-over-year. Adjusted EBITDA was $744 million, near the upper end of our previously communicated Q3 preview range, with an adjusted EBITDA margin of approximately 60%. Diluted earnings per share were $5.30 compared with $4.24 in the prior year period.

Turning to our balance sheet and capital allocation on Slide 8. We ended the quarter with approximately $1.5 billion in net cash remaining within our long-term target net cash range of $1.5 billion to $2 billion. Our balance sheet and projected cash generation continues to provide substantial financial flexibility.

In February, we laid out our approach to capital structure and capital allocation is prioritized maintaining the aforementioned $1.5 billion to $2 billion of net cash to provide resiliency to cyclicality and industry volatility. And to use additional cash to focus on growth technology deployment, innovation, R&D, technology road map advancements and M&A, especially related to complementary technology adjacent opportunities.

We further stated that we would evaluate applying cash generation in excess of the above capital priorities to share repurchases, pending clarity around third factors, including policy catalysts, new bookings volume and tax credit monetization. In line with this approach and following the Section 232 proclamation issued in Q3, our Board has authorized a share repurchase program of up to $500 million, effective immediately with no expiration date. This authorization reflects confidence in the long-term value of the business, while preserving our disciplined approach to capital allocation and ability to continue investing in technology development and growth.

Turning to Slide 9. Our full year 2026 guidance is updated to reflect a lower range of India domestic, international Series 6 and U.S. finishing volumes sold, with an associated reduction in revenue and adjusted EBITDA. As we stated in our original guidance in February, we entered the year assuming our India facility would flex production as needed to match successfully priced India-based demand in an inherently short book-to-bill cycle market.

We've accordingly reduced our full year India domestic volume sold forecast by approximately 0.7 gigawatts. We also stated in February that we entered the year assuming limited in-year demand for our Series 6 international product and bookings activity for 2026 deliveries has remained consistent with that assumption.

While the low end of our total U.S. component volume sold forecast reflects a modest reduction based on reduced Series 6 WIP share products sold, our forecast range for fully domestic products sold remains unchanged. As we look forward to next year, subject to a supportive policy and trade environment, we expect to increase exports from our India facility to the U.S. market to optimize the margin potential from this factory. And as Mark discussed, we expect to make a decision on the long-term operating path for our Malaysia facility in the first half 2027.

Core operating expenses primarily increased as a function of increased legal spend associated with defending our Topcon intellectual property position, which we believe has the potential to drive future revenue and earnings upside and increased R&D spend on perovskite development. Note, our full year perovskite related spend originally forecasted approximately $100 million is now expected to total approximately $120 million. This forecast increased in 2027 as we seek to advance this next-generation advanced technology.

Core OpEx increases were offset by a delay in certain start-up expenses of our South Carolina facility now expected in mid-2027, leaving our total 2026 operating expense guidance unchanged.

Net cash guidance remains unchanged. The reduced cash inflows from volumes sold, offset by a reduction to forecasted CapEx.

In summary, we remain well positioned as we close out 2026 and look forward to 2027, driven by U.S. policy framework supporting true domestic manufacturing, continued lower growth as the United States compete to lead in AI and utility scale solar ambition to meet the cost and speed to power requirements of a rapidly modernizing U.S. economy.

The drivers align well with our 44 gigawatts contracted backlog for mostly near-term deliveries and greater module supply availability in the 2028 beyond time frame, which we expect further improvements in pricing and contractual terms relative to our existing backlog. We continue to invest in our future, including our stage CuRe technology replication across the fleet, and increased perovskite spend in R&D, IP partnerships and our Perrysburg pilot line.

Our strong balance sheet allows us continued investment in the business while concurrently enabling capital return to shareholders.

And with that operator, please open the line for questions.

[Operator Instructions] Your first question comes from the line of Moses Sutton with BNP Paribas.

Maybe you could like elaborate on what seems like literally 0 urgency for developers to book '29 volumes. I sort of get the point on there's nimbyism there's positive economics, there's tariffs, but we also hear from developers they expect to really do a lot of safe harbor volume in [ 2029 ] Panel prices will structurally go up 50%. They don't have inventory that extends that far out, even if they do for the next 2 years. So -- sorry for the long question, but would there be a strategy change that's necessary? Would you cut price to $0.34, $0.35 and book the rest of the decade? It would seem that a little more agency for '29 than there used to be, maybe can elaborate a little more, because we do still expect at least 10 gigawatts of TAM each year in '29 and '30?

Yes. Well, look, I think the first thing I think we just need to maybe step back and reflect on is that everyone obviously is evaluating the execution against their development pipeline. And there's near-term initiatives and areas of focus that they're concentrated on right now and disruptions at certain times, as we indicated even as it relates to whether it's battery costs or other component costs, I mean, or the impact of tariffs that are swinging various projects around, people are all trying to work through that. They're trying to secure their financing. There's still interconnection challenges that people are working through. And then probably the biggest issue that people are continuing to struggle with right now is just the challenges that are continuing to happen in local communities even at the state level, in particular, from a government standpoint and trying to push back on deployment of data centers, as we indicated and maybe as a comment that we had in our prepared remarks.

So there's a lot that they got going on right now. So engagements that we are having right now with customers. And we are starting to book pipeline of opportunities that are going out into that horizon and some even beyond '29 going into '30 and beyond as well. But everyone is kind of like I got all these near-term things I'm focused on at this point in time. I've got to work through these, get everything addressed I'm still trying to digest the implications of what happens with the full implementation around 232. People are still trying to think through the Fiat guidance is going to come out sometime either later this year or beginning of next year. And so I think there's clarity that everybody is looking for at that point in time.

So I think it's just a matter of just the timing of letting some of this work its way through and where their near-term priorities are to then think through the balance of the decade and what they still need to do. I mean everyone is really concerned now meeting their 48 safe harbor requirements, which these products have to be safe harbored by and completed and commissioned by the end of 2028. And there's not as much focus going further out into the horizon, and that's largely what we're seeing from customers.

Just a couple of other comments. So we said on the call that we were seeing this year challenges in near-term book-to-bill on our international volume. And if that ties into what we said before, we don't want to be in the spot market. That's a challenging place for us to be. But if you look at our contracted backlog, we're still booked out a couple of years in advance. So when you talk about urgency around booking 2029, obviously, we'd like to see bookings momentum. We're open for business, but I don't think we have a desperate urgency to be booking in that horizon right now. And we also talked on the call in the prepared remarks around how we're thinking about contracting. And we're looking increasingly at individual projects versus large frameworks. And we're happy to take framework deals with the right set of structural protections within them. But we see more and more customers actually looking at individual project deals, and we've been looking at that. And generally speaking, that's going to give you shorter-term visibility relative to 2029. So if it's trying to book specific projects, that the amount of clarity you have for 2029 is a lot less than if you're looking at framework horizon. So there's that piece to factor in as well.

Yes. And one thing I should have mentioned, too, just I think it's just a really good example to highlight kind of the dynamics that are going on, right? Now we are in negotiations with the customer for 2 gigawatt procurement decision that was largely into final negotiations and documentation associated around the agreement and the particular state that this project was being developed put a moratorium in place. And the whole project stopped, right? And this has happened after the announcement of the 232. So customers are just uncertain and when that level of uncertainty does happen, then they're not going to be making commitments until they get better clarity.

Your next question comes from the line of Joseph Osha with Guggenheim.

Following a little bit on Moses' question. Alex, you mentioned that bookings we're seeing now reflect conversations that were undertaken prior to 232. Can you maybe give us a little bit of a preview on where conversations are landing for conversations that commence post 232?

Yes, maybe I'll add a little bit color around that. Since the announcement of 232, we have revised and increased our pricing. In some cases, they were deals that were still in the early stage of negotiations. So we have increased pricing in those negotiations with customers. It -- and if it's a new opportunity that's come in our way that wasn't even in any form of conversation prior to the 232, then those prices have gone up relative to what we would have been indicating previously.

We still have to work through all those issues, right? And I think we're -- anytime you get through here in an environment of rising prices, there's a lot of price discovery that's going to get done. Everyone want to go out and they want to evaluate all other options and sort of discover where they think the pricing can be for different dynamics and whether they need the domestic content or how much of the domestic content from a point standpoint do they need. So they're going to do a lot of heavy lifting to sort of figure out and triangulate what makes the most sense.

Partly because they also have to, in most cases, go back and get approval through their boards. And when you see a price point in the market that was there previously and now that, that price point has increased significantly, then everybody has to then sort of do the due diligence around that. So what I would say is the deals that -- some of the deals that we have announced in this quarter were priced before the 232 in cases where we were in earlier negotiations with customers after the 232, we have looked to try to increase that and then clearly, any new opportunities that kind of informed the more comprehensive view of what we think the pricing environment is like today. But like I said, the price discovery is going to take a while before you'll start to see that volume being executed.

Your next question comes to the line of Philip Shen with ROTH Capital.

In terms of price, just to continue with the conversation, the price discovery may take some time. It sounds like you are -- you have 4 gigawatts, of which maybe 2 gigawatts needs to be decided on in the next 5 months. And so just curious where -- what are we talking about now in terms of where pricing could go to in the crystalline silicon world, we've quantified it as pricing has gone up $0.10 to $0.20 a lot depending on the product and maybe timing. And so that's a big move, and I understand the price discovery patients or timing that people need because they have to process kind of this upside shock to their economics. And so I was wondering if you can share with us a little bit of where do you think -- what could you book in the next quarter and then what kind of price do you think you might get?

Yes. Phil, the one thing on price, too, I just want to make sure that we've highlighted this before, it does vary by -- we have it in our prepared remarks as well. I mean the price for us -- for our Series 6 product is different than our Series 7 products. Our domestic -- high content domestic content product is different than a fully international imported module. So there's going to be a distribution of what that pricing could look like. And so if I'm selling something that's being imported from Malaysia as an example. That's going to be at the low end of that continuum. And if I'm selling our Series 7 product that's domestically manufactured here in the U.S., and that's going to be at the high end. So there's going to be a continuum around pricing.

As it relates to -- we just announced, again, I think, another quarter at $0.36 or something like that is what our -- since the last call, Con call, I think it was around $0.36. And we have deals and that's an average. We've got deals that are above that, and that's a momentum that we'd like to continue to carry on -- carry forward. But I also want to be careful of we've got to premarker out there with so much that still is as fluid and still evolving. I don't want to give an indication of what that price realization could be. We're just going to work to try to see how we can engage the market and capture the highest value relative to competitive alternatives. And Chris and silicon stays in the range that you referenced and then that gives us quite a bit of opportunity to improve pricing dynamics for our technology. But a lot of work still to go to get there.

So maybe just two other comments. As it relates to the volume that we already have that may get elected by customers, that's going to range from somewhere close to current backlog average or somewhere close to where we're pricing now. So you're going to see a variety in that. But it's going to be between what we see historically and what we're seeing currently.

And then as you think about future pricing, Christmas silicon pricing is still most likely for deals in the next 12 to 24 months. And I don't think there's a good market for that price point when you get out into the late '28, 2029, 2030 time frame, which is when we have most available product to sell.

So one of the reasons I think customers are still doing price discovery is they don't necessarily have a good benchmark that far out, that's reliable or trustworthy. And so we're still working through our pricing relative to where we think the market is and customers trying to understand our pricing relative to where they think that market point will be.

Yes. And I think the other point maybe I'll highlight as well, Phil, is that, look, we still have a little bit of time before we get to full implementation on the 232. And what I will say is, I'm very happy to hear the kind of enforcement focus that the commerce has to address stockpiling. The problem is there's such an aggressive posture and risk-taking that people are willing to do in order to get a module into the U.S. at some pre-32 MIP. And so we're in a lot of stories of to what extent people are going to in order to try to get product into the market, whether it's a set the cell level or whether it's at the module level and just aggressively pushing the envelope. And I know Commerce is doing a great job of trying to enforce that and to prevent that from happening and stopping that. But the onsite of this stuff is coming from every angle. .

And so I think there's also that piece that people are waiting for. Let's let -- maybe there'll be a lot more product that does make it into the market somehow. And a customer may say, I'm willing to take some we buy some time because maybe I can get an opportunity to get some pre-32 volume that's been imported into the market at some increasingly low price. And that's informing their decision or at least create a willingness to wait, and we obviously are there to encourage commerce to do whatever they can to ensure that enforcement of preventing that stock filing. But that's not stopping the activities. People are willing to take a lot of risk, and it's very aggressive, and we'll have to see how effective commerce could be. So far, we applaud them for their efforts, there's no doubt about that.

Yes. So far, they blocked , I think we've learned about 60 importers blocked that up until about 2 weeks ago, and that's out of an importer count of, I think, 400 -- pre-232, we learned that 200. And then post-232, that jumped up to 400. They've locked at least close to 60. So it looks like they're working -- the enforcement is pretty aggressive. Any thoughts on that? And then how much higher MIPs could go up and what kind of timing that could be?

Yes. I think on the enforcement, I mean, I -- some of you may know that we did our ribbon-cutting for our new factory in South Carolina and secondary Nick was there and a number of the key members of his team were there. So it gives us a great opportunity to talk with them and to sort of continue to emphasize the importance of this enforcement, and they're 100% committed on the enforcement side and to do everything they can to prevent those types of things from happening.

And where we can, we are sharing insights with them and things we think could be helpful from that standpoint. And just to your point, I mean, the risk tolerance and that people are willing to take knowing that there's going to be that level of enforcement is not slowing some things down very much. I mean, yes, you can get 60 out of the -- if the number of importers of record went up from 200 to 400. And it's great that we caught 60 of them, but there remains another 140 that potentially are still getting around the current enforcement levels, right?

So -- and then as it relates to the MIP, look, I think is very clear if you haven't seen -- remarks at our ribbon cutting, I think you just ought to watch it. There's a couple of videos that are out there, and he's 100% clear the strategic intent of what we're trying to do here, which is to create U.S. manufacturing, U.S. jobs, U.S. supply chain, U.S. technology and things that can help enable energy independence and national security.

And so they're going to enforce this and they're going to put the right requirements in place of ensuring the domestic manufacturing can thrive and then put as many incentives as possible to encourage people to manufacture more here in the U.S., which I think is great.

Your next question comes from the line of Brian Lee with Goldman Sachs.

I know we're kind of dancing around the same topic, so I'll try my best as well. But it sounds like you guys are moving from being a contract go-to-market strategy to more of a spot market go-to-market strategy, which I think most of your peers do and now is the case even for First Solar a few cycles ago. So maybe, as you said, Mark, the bookings cadence moderates here with all the uncertainties, but the bookings cadence can potentially be more fluid on a quarter-to-quarter basis. I guess, is that fair? And are you equipped, I guess, from a sales force perspective to be able to manage through that?

And then again, on the pricing entitlement side of things that you keep referring to developers and all the uncertainties that they're having to grapple with. I guess what are you tasking your sales force with -- in terms of this price discovery process post 232? I would imagine they're not going to go anywhere lower than what you've been looking at recently, which is $0.36. But -- is it a range of like $0.36 to $0.40, $0.40 plus? Like what are the puts and takes around kind of what they're going out to market with?

Yes. So I think what we were trying to say, Brian, just as it relates to, a lot of the deals that we announced going back 3 or 4 years ago when we built our contract backlog. A lot of these were multiyear framework agreements that have indicative delivery schedules associated with them. And then as we said before, they had certain rights -- annual ship rights or quarterly ship rights, things could move around from that contracted schedule, but it wasn't tied to a specific project.

And what's happened when we do that is the volatility of our customers' development pipeline for various reasons is just whip stuff all over the place, right, in terms of, hey, this is what the contract was -- and now this is what the current requirements are given the state of their development portfolio of projects and what have you. And what we're trying to do now a little bit and not that we won't get back to that, we've got a couple of very large deals where there'd be more of that legacy kind of framework agreement, strategic partnerships. And we'll clearly do those. But near term, what we're seeing more is that everybody is like, I got a particular project, and I want to negotiate this particular project. And that gives us better clarity and uncertainty of the project where it sits in its permitting process where it sits in its interconnection process where it is in its financing, if it's approaching that or at least we know what the time line and the profile looks like and gives us a better clarity of the timing of when the shipments need to occur. So that's where the near-term focus is on that. But what I would say is that after this price discovery kind of it's better understood. I think we may see some customers moving into that other direction to say, "I really want to lock in and secure my project pipeline going out into '29-'30. And I said, we have some customers that are even wanting to start having some conversations that go beyond 2030.

So as it relates to the sales force, I somewhat alluded to that, first off, from a capacity standpoint, I don't think it's an issue with our capabilities of our sales team to handle that pivot, right, to more product-specific versus framework structures. But as I somewhat alluded to, is we have gone and we've repriced deals that were in the early negotiation phase and any deals that have come to us now are obviously reflected of what we believe the current market pricing should be.

So I'm not going to give you an indication of what that looks like. But we have recalibrated and I think the market fully understands that. But I also want to continue to go back. I mean there is a continuum of what those prices are going to be in near term, and I think if you alluded to this in prior calls, what we have right now available to book is more Malaysia finished good module coming into the U.S. or an India product coming into the U.S. And so that profile in terms of ASP entitlement is going to be on the lower end of that continuum, as I indicated, the lowest price point we would have for a module will be for that international product coming fully finished international product coming into the U.S. and then the highest will be kind of 100% domestically manufactured Series 7. So it can move around that average is reflected in the numbers that we've been giving you. But in any particular quarter, if I'm heavy on maybe international modules coming to the U.S., that would drive that average ASP down. And clearly, we'll provide color around that. But yes, I mean, pricing is reestablishing itself at a higher point than it has been historically, and we'll continue to see how price discovery works through

Brian, just one follow-on point to that framework discussion. And 1 of the challenges that we've talked about over the last couple of years now is the period cost that we've been incurring around the amount of inventory we have to keep, the storage cost at a entry overall sales rate. If you think about when there's an ability to move volume around or when customers need to do that because the timing of their projects shift. I have to keep inventories to make sure I can see is all the available contracts for the house. That generally means my -- at my overall inventory position as well as the working capital position is worse.

At the same time then, if that product all does move to the right from a schedule perspective, now I end up having the warehouse materially more volume than I expected. So generally speaking, it's become harder for us to plan around that, and we're seeing a significant financial impact from that. if we move to more of a project-based model where there's a certainty around an individual project, the timing of that project and delivery schedule that allows me to plan on the operational side, much more efficiency, reduce working capital and reduce storage costs as well. So provides that operational financial efficiency that we're looking for.

Your next question comes from the line of Jon Windham with UBS.

Perfect. Mark, I appreciate you taking the questions. We'll get on to the exciting topic of filing your taxes. So in the presentation, you note there's $1.9 billion of government grant receivables. My understanding is that's payable 2 to 3 months after you file your annual returns. Can you just give us an indication of how your planning is for that cash actually going to your balance sheet? -- and then into your bank account? And then any comments around the liquidity in the secondhand market if you were to decide to try to monetize those early?

Yes. So you're right. On the balance sheet, there's about $1.9 billion. About $100 million of that is related to India. So you got about $1.8 billion in the U.S. And then part of that $1.8 billion relates to credits that were generated last year and we have yet to file -- yet we have just filed our tax return we get to receive refundability for those. But if you look at this year's credit generation, the forecast is about $2.1 billion of credit to be generated this year. If you look at the 2024 credits, we sold most of those. We did keep a portion of them to file our 2025 tax return in October of 2025, and we received refund for what was about $120 million in about 3 months' time.

And importantly, even within that 3 months, the government was actually shut down for about a month of that. So the turnaround time from filing to receiving the cash relatively short. As it relates to this year, we haven't made a decision yet. We've not made any assumption around a reduction in the carrying value of those credits. If we were to sell them. We've seen that the market through the first half of this year, I think, was a little softer to credit sale value. That's picked up pretty significantly towards the back end of the year as large corporates have better position and understanding of the tax position coming into year-end and looking for tax planning actually. So we're seeing a lot more how reach and interest around those credits. We're in some discussions around potential sales there. But we'll look to balance how much we would like to sell to pull some liquidity forward versus how much we're comfortable filing for fundability for because then you're not taking a discount on that sale. So it's a balancing after we're in discussions around those as well.

Your next question comes from the line of David Arcaro with Morgan Stanley.

I was wondering with the shift towards individual project booking, typically, what's the time frame for how far in advance you would end up getting that order into your backlog relative to, I guess, the timing of the project? And I think you had suggested that the decision on Malaysia could be made in the first half of 2027. Are you thinking that we'll get enough kind of visibility by that time that you'll be able to make that call on what to do with that capacity?

Yes, I'll take the first one and I'll let Alex take the Malaysian comment. So yes, so basically, what happens on a -- if we're being specific to a project. Again, modules use at the forefront of your procurement-led decision, you want to secure that technology because it also drives your design of the factory, right, or the sign of the project, excuse me. So crystalline silicon and Catell design your factory and your plant, excuse me differently. There's different string count, different voltage and other parameters, which you would want to optimize for. Generally, we don't have the same constraints as it relates to the tracking -- backtracking requirements because we don't deal with the same efforts impact of shade or defuse light and those types of things. So there's a lot that has to be done early on. And so you're going to want to source that module and lock that up so you can be unencumbered now to figure out do you want to design and optimize at the project level.

So it could be 18 months, 2 years before the first delivery would start to happen or at least for sure until the project would be commissioned. So it doesn't really -- it shortens your time line a little bit from a visibility standpoint, but it doesn't really significantly shorten it. But it gives you better certainty. -- there's no doubt about that. People would be willing to go out into 2030 in a framework kind of construct. And projects normally would be commitments that would go out into 28 or into early '29 is probably not as far out as they would want to go, but still more than enough visibility to properly understand the project and price it appropriately.

And David, as it relates to Malaysia, yes, I do believe we have enough visibility going into the first half of next year. We've effectively been carrying that asset running at very low throughput for the whole of this year. And we've done that deliberately, we could thought the and the utilization costs associated with that was worth the option value of having the ability to run what would be up to about 1.8 gigawatts of international volume.

As we look through the 232 announcement, as Mark said, the customer base the transition fully understanding what that looks like and trying to work through what their need is going to be. We want to make sure we have the optionality in Malaysia in the event of that significant demand does materialize. We've also said that we have an order of how we want to make sure our factories sell. So first, we look to make sure our U.S. content is sold. Secondly, we look to make sure we're selling our finishing line production of our utility in South Carolina. And then after that, the full international product. We're also going to be bringing more volume from India over into the U.S. next year, subject to the trade and tariff environment that makes sense. So I think we'll have the understanding of the marketplace that we'll need by the first half of next year, and we're still looking at that as an option value and seeing how the effect of the 232 plays out.

Your next question comes from the line of Vikram Bagri with Citi.

Apologize for another question on pricing, but this is important. On this call itself, you've heard pricing either go down by a couple of cents or go up by $0.20. So Mark fully appreciate the lack of clarity about the environment we are in. But for CD7, is there a construct you can provide for pricing that we should keep in mind. So if pricing for silicon-based panels move higher, we should assume a certain discount to that, which some people will characterize as quality discount.

And I imagine pricing also depends on several of the factors, project and contractors, but all else being equal, is there a construct where pricing for silicon-based panels moved higher than $0.47 or $0.45, we should assume a certain fixed discount for CD7 -- UA CD7. And then related to that, Mark, you talked about a number of items, limiting the ability of developers to make long-term plans. Is competition in panel manufacturing is also one of the factors, which is sort of like prohibiting them from committing to supply of panels. Do you think the lack of urgency to book 2029 is also a function of new capacity coming online potentially in the U.S. from other suppliers of panels? Or how do you see the competition evolve over the next couple of years in the sector?

Yes. So I'll talk on the -- so if you take S7 as an example, -- just one thing to make clear, you made a comment about quality discount. I don't know if you're familiar with how energy predictions are done for crystalline silicon technology, they actually have a quality factor that they use, which would basically reduces the energy profile of the technology as part of the energy prediction on which that does not apply at all to First Solar's technology. So we don't have any quality factors that are assumed in the energy predictions that are underwritten by and supported by the -- the banks. So I'm not sure what your point is around that, but crystalline silicon clearly does. They have a meaningful quality factor that is being used when they do the energy prediction for that technology.

What I would say, if you take a Series 7 CuRe product and you compare it from an energy profile because again, energy to what our customers pay for. And what the value of that product would be relative to a PERC technology, it would be an advantage to PERC, so there would be no discount to PERC. It would also vary with where it is because obviously, our technology does not have what CuRe will have better bifaciality than our QED product is our bifaciality is lower than crystalline silicon. So if you were to go further north East where you get a better impact of bifaciality, then you'll see a different outcome.

If you go to the other Southwest or to the Southeast and more hot and human hot climates or hot human climates, you're going to see a better outcome for our technology. And then if you compare us to Topcon, which again, the vast majority of anyone producing or selling a Topcon products in the U.S. is a prince on our IP, which we will stand by and completely enforce that, then you'll have a slightly different outcome. We would actually be average slightly disadvantaged to Topcon of a few pays. So it depends. I mean you've got different situations, different geographies, different terrains. A lot of different things factor into that and determine the relative competitiveness of the technology.

So -- but I would say general rule thumb, Series 7 going to be advantaged against PERC and disadvantaged by relative to Topcon.

As it relates to the competition and how that all plays out, you have to think through how much are you willing to -- first off, is the competition in the manufacturing that's coming in the U.S., truly Fc-free, truly unencumbered by any dependency from the Chinese, whether it's the supply chain, whether it's the tools and the manufacturing equipment, whether it's the IP, right?

A lot of our customers do their own subjective analysis and have concluded that there is a substantial risk still that these entities are not Fc-free, and they don't want to take on that risk. That puts us in a much better position. There's no doubt about that, and that's also why I think you're going to find customers leaning into developing partnerships with First Solar that go longer dated. I believe I remember reading one of the filings recently for one of the U.S. manufacturers, and I believe almost 100% of the sales that they're generating this year is to a related party, and I understand that related party to be a Chinese counterparty.

So I mean it tells you just looking at that as an example, that in certain situations, customers are not willing to take on that risk. And we still don't have full clarity around FAC yet, and I think that's why people are being a little gun shy.

Our final question comes from the line of Dimple Gusai with Piper Sandler.

My question is regarding just December 4 in terms of outcomes and what you're hearing. I mean there's been a lot of news recently, T1 recently filed for a waiver around cell imports. We're hearing a lot about lobbying for exclusions. And we also know utilities are pretty price sensitive here. So are we -- in your view, is there a risk to the upside or downside just in terms of the outcome there. And then separately, just kind of talking about some of the news that we're hearing about is also news on Swift Current low suit that's been in the news. Is there anything you can speak about here? Is there any merit to that? Is there a size associated with it? And any merit really?

Yes. As it relates to the exclusions, look, I do believe because, again, the program incent and motivation is to try to ensure manufacturing here, reshoring of manufacturing, true manufacturing, truly unencumbered by Chinese or adversarial countries dependencies. Each one of those will be determined on a circumstances basis and work through with commerce to say is their true substance to this. And then in general, the focus with commerce is to limit the period of time at which that exclusion would be provided.

I don't stay on top of it. I don't know. My team may have better insights than I do in terms of how any exclusion been granted at this point in time. It's not clear to me that there have been any granted exclusions. I know there's been some representation of statements that have been made, but I don't believe there's any actual exclusions that have been communicated publicly at this point in time. But that's something that we're going to have to, to stay close to and understand.

I'm perfectly fine though competing. We've always said U.S. manufacturing on a level playing field First Solar is more than happy to compete on our own merits and our technology and the competitiveness of our manufacturing process in our supply chain. So some of that does happen. And then it will evolve, and that will create the domestic industry that I do believe still is a strategic thread that's important for us as a country.

Assuming we do it right, right, assuming it truly is U.S. technology, U.S. manufacture and U.S. supply chains, right? Creating American, U.S. jobs and supporting American families. There's no doubt about that. That's absolutely critical from that standpoint.

As it relates to the lawsuit, I really can't comment much of anything around that. We actually will be filing our response here shortly for the required time line. And I think there'll be better clarity around the issues in the allegations that have been raised during that -- during the -- part of that lawsuit. What I will say is that we have always -- even with the engagement with Swift as we've always said we will stand by and we will support under the parameters of our limited liability warranty, the obligations that we have. Again, that largely is reflective and relates to the Series 7 initial product launch that we had indicated that there was a couple of issues that came out as part of that launch that resulted in an initial stabilization of that technology to be below expectations.

And then underneath that program we put in place. We said we would remedy those issues with customers. And we engaged even with Swift to have that have that conversation. I think where the challenge comes in is where people are asking for revenues that are outside the parameters of the warranty. And that's just -- that's not something that we will do. But we will be more than -- we'll stand behind the technology, and we have that conversation with Swift in this case. And they chose to try to pursue a different approach where they think they could be successful. And again, you should see our response in our filing coming out here in the next several weeks.

We have reached the end of the Q&A session. This concludes today's call. Thank you for attending. You may now disconnect.

Good afternoon, and welcome to First Solar's Second Quarter 2026 Earnings Conference Call. This call is being webcast live on the Investors section of First Solar's website at investor.firstsolar.com. [Operator Instructions] And please note that today's call is being recorded.

I would now like to turn the conference over to your host, Byron Jeffers, Head of Investor Relations.

Good afternoon, and thank you for joining First Solar's Second Quarter 2026 Earnings Call. With me today are Mark Widmar, Chief Executive Officer; and Alex Bradley, Chief Financial Officer. Mark will begin with second quarter highlights, followed by Alex, and then we'll open the line for questions.

Today's discussion contains forward-looking statements. Actual results may differ materially due to risks and uncertainties as described in our earnings press release and other SEC filings and the earnings material available at investor.firstsolar.com. We undertake no obligation to update these statements due to new information or future events.

We will also reference certain non-GAAP financial measures. Reconciliations to the most directly comparable GAAP measures are in our earnings press release and presentation. This non-GAAP financial information is not intended to be considered in isolation or as a substitute for financial information presented in accordance with U.S. GAAP.

With that, I will turn it over to Mark.

Thank you, and good afternoon. Beginning on Slide 4. We delivered both record second quarter and first half sales volume and improved financial performance relative to the prior year. During the quarter, we generated over $1 billion in net sales, expanded gross margin to approximately 57% and delivered strong adjusted EBITDA performance.

We also surpassed an important milestone for First Solar, exceeding 100 gigawatts of cumulative module sales globally. We view this as a reflection of the trust customers have placed in First Solar over the more than 2.5 decades and the durability of our technology and manufacturing platform. We ended the quarter with approximately 45.1 gigawatts of contracted backlog. We delivered -- with deliveries extending through the end of the decade, demonstrating the demand for our differentiated technology platform, domestic manufacturing footprint and delivery certainty.

Turning to manufacturing. Our U.S. facilities continue to operate at high utilization rates during the quarter. In South Carolina, the first phase of the finishing facility remains on track to begin production in the second half of 2026, with equipment installations progressing as expected. For the second phase, we now expect completion in mid-2027. While the revised time reflects a number of factors associated with optimizing the facilities launch, it also enables the earlier incorporation of CuRe technology.

We are pleased with the performance of CuRe with both high-volume manufacturing at our Perrysburg facility and performance data from field deployments across multiple climates exceeding expectations. We believe incorporating the technology closer to the onset of the facility's commercial launch will simplify execution, accelerate value realization and enhance customer value and the facility's long-term financial performance. Once completed, the South Carolina facility is expected to provide up to 3.5 gigawatts of finishing capacity for modules initiated at our international manufacturing sites, giving us greater flexibility to optimize our supply chain flexibility while also optimizing freight, tariffs, domestic content and Section 45X economics.

With respect to our international manufacturing fleet, production planning and utilization levels in Malaysia and Vietnam continue to be influenced by U.S. market demand drivers and economics, including the pending Section 232 polysilicon and derivative investigation and tariffs. We expect greater policy clarity will help inform the long-term operating profile for the approximately 1.8 gigawatts of fully finished international capacity that remains available after accounting for capacity being used to produce semi-finished product destined for our new South Carolina finishing line.

A note on manufacturing optimization and allocation. Approximately 41 gigawatts of our 45 gigawatt backlog includes some form of domestic content requirements. These requirements vary significantly and range from requiring exclusive supply from U.S. fully integrated factories to blending U.S.-made supply with both fully integrated domestic factories as well as product from our upcoming South Carolina finishing line to a domestic content points requirement, which is factory-agnostic, allowing blending of product from across our global fleet.

We therefore continually balance and refine our module supply and demand allocation across the fleet to meet customer contractual obligations, optimize factory throughput and optimize gross margin. This typically means that over a period of time, we will seek to maximize production and sales firstly, from our fully integrated U.S. factories; secondly, from our South Carolina finishing line; and thirdly, from our international facilities.

As it relates to prospects, we continue to advance our development program for this potentially significant technology platform. Our previously announced development line continues to progress -- to process improved efficiency and reliability attributes on smaller form factor modules, while our Series 6 form factor pilot line remains on schedule and is expected to reach operational readiness in the first half of 2027. Our continued progress has given us confidence as we continue to invest substantial capital in our efforts to realize the commercialization of prospects.

Earlier today, we published our latest corporate responsibility report, reinforcing our conviction that how and where solar technology is made matters. The report details how we create enduring value by developing, sourcing, manufacturing and recycling solar modules domestically, supporting jobs and communities, strengthening industrial capacity and help ensure the benefits are realized locally. It also highlights our continued focus on responsible manufacturing, supply chain transparency, workforce development and resource efficiency. The report reflects the effectiveness of a business model where [ corp ] responsibility isn't a construct, but the default.

Before turning the call over to Alex, I want to briefly address the market and policy environment and how it is informing our commercial approach. The underlying drivers for utility-scale solar remain intact, including load growth, data center development, electrification, aging generation assets and the need for affordable, scalable new capacity. The policy landscape continues to evolve, particularly as it relates to pending outcome for the Section 232 polysilicon and derivatives investigation as well as final FEOC regulations.

In this environment, we continue to prioritize pricing, contract quality, appropriate risk allocation and long-term value over short-term bookings volume. Relative to the beginning of the year, we are seeing increased customer engagement. And as policy clarity improves, we believe First Solar remains well positioned to capitalize on these opportunities.

With that, I'll now turn the call over to Alex to discuss our bookings, financial results and outlook.

Thanks, Mark. Beginning on Slide 5. As of June 30, 2026, our contracted backlog totaled 45.1 gigawatts with an aggregate transaction value of $13.6 billion, exclusive of technology adjusters, with scheduled deliveries extending through 2030.

Early this month, Cypress Creek Energy broke ground on the Steel River Energy Center in Arkansas, a project utilizing First Solar modules and previously included in our contracted backlog. The initial phase is expected to provide approximately 1.6 gigawatts of solar generation capacity and 1.9 gigawatt hours of battery storage to support Google's growing energy needs with the opportunity for future expansion.

Since our last earnings call, we recorded approximately 1.9 gigawatts of additional U.S. gross bookings at an average selling price of approximately $0.36 per watt, inclusive of applicable technology adjusters. While near-term customer activity continues to be influenced by the current policy environment discussed by Mark, our fully integrated domestic manufacturing fleet remains substantially committed through 2028, providing a high degree of volume and pricing visibility. Given the limited amount of uncommitted domestic capacity available over the next several years, we continue to be disciplined in evaluating incremental contracting opportunities.

We also initiated our first customer notifications related to contractual CuRe adjusters during the quarter, an important milestone in beginning to translate CuRe's performance benefits from potential ASP adjusters into backlog value and future revenue realization. We expect the contribution from these adjusters to increase as CuRe deployment expands across our contracted portfolio. As a reminder, we expect limited ASP upside from CuRe sales in 2026, largely as a function of contractual notification deadlines relative to the timing of the decision to recommence CuRe production.

Turning to India. Our guidance continues to assume production is largely sold domestically in a short-cycle book-and-bill market, with the factory operating at high utilization rates. India gross bookings during the first half of the year totaled approximately 1.1 gigawatts and average selling price of approximately $0.20 per watt. Given the shorter contracting cycle of the domestic Indian market, booking economics generally provide a reasonable indicator of near-term revenue realization, subject to [ no foreign currency book ].

Turning to Slide 6. Net sales for the second quarter were approximately $1.06 billion, a decrease of approximately 4% year-over-year. The decrease was primarily driven by lower revenue associated with customer contract terminations recognized in the prior year period, partially offset by higher module volumes sold. Gross margin was approximately 57%, an increase of approximately 12 percentage points compared to the second quarter of 2025. The increase was primarily driven by an estimated $89 million net IEEPA tariff-related benefit, a higher mix of modules qualifying for Section 45X tax credits and lower logistics costs.

The net IEEPA per tariff-related benefit reflects our current estimate of expected recoveries related to commercial obligations and other tariff-related considerations and remains subject to refinement as additional information becomes available. These benefits were partially offset by lower termination related revenue and higher duties and tariffs.

While logistics costs improved year-over-year, the quarter includes higher over-the-road freight costs driven by overall capacity tightening and volatility in diesel costs. These impacts were partially offset by higher sales rate recovery. Operating expenses were approximately $155 million, including $76 million of R&D expense. R&D increased year-over-year, primarily reflecting continued investment in [ Propco ] development, the impairment of certain R&D equipment that is no longer expected to be used as part of our technology road map.

Net income was $423 million, up approximately 24% year-over-year. Adjusted EBITDA was $644 million, above the high end of our previously communicated Q2 preview range, with an adjusted EBITDA margin of 61%.

Moving to Slide 7. We ended the quarter with approximately $1.7 billion of net cash, providing substantial balance sheet strength and financial flexibility while remaining within our targeted long-term cash range of $1.5 billion to $2 billion. Operating cash outflows year-to-date were $360 million, reflecting first half working capital dynamics and improved compared to outflows of $458 million during the first half of 2025. First half capital expenditures were $280 million, primarily supporting our South Carolina finishing facility and technology investments. We completed the full prepayment of our India DFC loan during the quarter.

Turning to Slide 8. Our full year 2026 guidance remains unchanged. With that said, our guidance now assumes a net tariff impact of $60 million to $80 million, with updates including the previously mentioned net IEEPA recovery and the assumption of Section 301 tariffs in the second half of the year. We also forecast offsetting updates between production start-up expense and R&D expense, as well as incremental freight costs due to certain nonrecoverable domestic trade expenses above our previously assumed forecast, largely driven by changes in module delivery locations. And note, in some cases, domestic freight costs are now approaching international shipping economics.

For the third quarter, we expect volumes sold between 3.9 and 4.5 gigawatts and adjusted EBITDA between $625 million and $775 million. In summary, our first half performance and reaffirmed outlook reflect the strength of our strategy of reshoring and scaling domestic manufacturing, progressing our technology road map and maintaining a selective approach to new bookings in light of key pending trade and policy dissemination. As we look ahead, our priorities remain unchanged. We remain focused on disciplined execution, serving our customers, advancing our technology road map, managing capital prudently and maintaining financial flexibility.

And with that, operator, please open the line for questions.

[Operator Instructions] Your first question comes from the line of Jon Windham with UBS.

So obviously, the SEC had a ruling about solar inverters a couple of days ago. And I think on one side, it goes along [ the shaft ] serious the government is in promoting domestic content within -- especially electrical equipment, hardware, which is obviously very good for you given your position in domestic solar modules. But just curious if you have any early thoughts on potential impact on broader solar installations and the ability to work around the industry to work around that provision? Thank you so much. .

Yes. Thanks, Jon. I think if it continues the same of -- our U.S. government trying to ensure that we don't have any overreliance on that serial countries. And obviously, China being one of them in particular. I think the good thing about this is that the industry has started to get ahead of trying to find domestic supply chains, comprehensive domestic supply chains. We obviously were an early industry leader in that regard of reshoring manufacturing and creating a supply chain here in the U.S. for our U.S. operations. You're seeing this now really across all components of equipment suppliers, all the way up even to try to find localization for the battery supply chain as much as you can.

So I don't see it being a constraint near term. I think the current models that [ and the shipping ] into the U.S. will continue to be allowed to be shipped into the U.S. I do think there is a theme or a message there, though, that, that scrutiny may be stepped up as we move forward. But I think it just sends another great signal to domestic manufacturers that look, we need [ to before ], we need domestic supply chain resiliency to enable not only the solar industry aside, but really all of the industries that as we reindustrialize in the U.S. economy, right? So again, I think it's a good indicator of a continuing theme and message that this administration has and we fully support it.

Your next question comes from the line of Brian Lee with Goldman Sachs & Co.

I just had two. I guess, first, on this Google-Steel River project, I appreciate you guys commenting on that. I might have missed it, but how much of the 1.9 gigawatts in U.S. gross bookings came from that 1 project in the quarter? And then how much more bookings potential exists on that project side?

And then your bigger picture, maybe speak to how you're seeing general interest from that hyperscaler data center community? And then second question I have is just kind of the customary latest thoughts, timing, visibility into Section 232, how you're viewing the potential for floor prices in the $0.40 per watt or higher range? And then how quickly do you move on your bookings funnel in Southeast Asia strategy once you get clarity on this presumably, hopefully in the next few months?

All right, Brian, I'll try to take kind of the first two, and then Alex will talk maybe a little bit about the views of Southeast Asia. So make sure it's clear on the project that we announced with our partner that we supplied models to for Cypress Creek, that is already in our bookings, okay? So that was just a highlight. It's a great project.

If you actually look at some of the more recent announcements that have been made over the last several weeks, I think you kind of see a theme there. You've got a very large project with Cypress, the one that we've referenced that it will be Phase 1 of kind of call it the 1.6 gigawatts, then it goes to Phase 2, which we believe is about 2.5 gigs. So that's a very large project. And I think the battery component of that as well is going to be north of 2 gigawatts of -- megawatt hours from a battery standpoint.

Really important strategic project. It's there to support Google. We have 2 other projects that have been announced over the last couple of weeks, 1 with Terra-Gen, which was about 1.4 gigawatts. And then we had another 1 with [ Panama ], which was another gig plus. So those 3 projects that have been announced recently are about 5 gigawatts of capacity. The [ Panama ] -- part of the [ Panama ] volume was actually announced last quarter. So when we did the announcement last week -- last quarter around bookings volumes, which I think we had in total around [ 1.4 ], [ Panama ] was actually included in that volume.

But I think it's a great message that the demands there, half of that volume of that 5 gigawatts I referenced is directly communicated and tied to Google as a hyperscaler. The other 2.5 gigs, I haven't disclosed the counterparties. But if you look at the verbiage around the announcements on that, they'll reference a very large corporate account, 1 of the largest companies in the U.S. You can kind of get a sense of the likelihood of who that counterparty is going to be for that project.

So strong demand for continued demand for hyperscalers, really strong relationships and partnerships with First Solar to support those types of strategic projects that really kind of thrive on the importance of certainty. Right? Those projects are strategic. They're important. They obviously include stores as reflected in the Cypress Creek project.

As I've always said, the first thing you need to do is you're building out your projects and derisking is that you need to make sure that you have a reliable partner who can make sure the photons become electrons. Without that, the whole project is going to sit at risk. And we can deliver that certainty and that great technology and that reliability. So we're seeing that in the marketplace and continued strong interest, but driven by -- as currently still some [indiscernible] demand from hyperscalers.

As it relates to 232, I'll take the pricing piece and then Alex will talk to kind of how we throw that into our views around Southeast Asia. Look, it's still -- there's still a lot of views out there. I think everybody has a view of how the construct may be with minimum import price and then maybe with the tariff on top of that, there's some views of whether there's quarters or not.

All I can say is still evolving. And we do believe it will be constructive. I don't want to give kind of our internal read of what we think it potentially could be because there's still a lot of open pieces. I can say that we're still in constant contact with the appropriate parties at USTR and commerce to continue to bring our voice into the conversation. And we're still optimistic that the outcome will be constructive. And we've used it as a reason to be disciplined, and we'll see what happens once it's finally announced.

And there's demand that's still sitting there on the sidelines. If you look at our cadence and our momentum around our bookings, just here in the month of July, we booked almost 2 gigawatts in the U.S. at very good prices, as Alex indicated. There's about 2 more gigawatts, north of 2 gigawatts, it sits into a contracted subject to CP. And then I've got another 2 gigawatts of active conversations with customers that there's a high probability we can close through by the end of the year. So -- and we'll see how much that gets further catalyzed by decision around 232.

Brian, as it relates to [ sales days ] capacity, we talked on the last couple of calls around looking at this a bit like an option. So we're running somewhere around $30 million quarter of underutilization associated with running Southeast Asia manufacturing well below its theoretical capacity, about half of that cash, about half noncash. Given that we've been holding through the first half of the year, making a decision on the long-term future there pending the outcome of 232, it makes sense to continue to do that. So I still view this as we're waiting for the outcome of that policy.

And just to frame the amount, if you were to go back and look at the slides we put out in our February call, it shows you nameplate capacity of production. So we originally had about 7 gigawatts of total capacity sitting in Malaysia, Vietnam. About half of that is going to be dedicated to production that will feed our new finishing line in South Carolina. So there's about 3.5 gigawatts left. Of that, we did take out some tools, bring them over to the U.S. to reuse in our [ Brodsky ] work.

So ultimately, it leaves us with about 1.8 gigawatts of end-to-end fully finished capacity that we could ramp up across Malaysia, Vietnam. So it's about that 1.8 that we're talking about. We're thinking -- we're holding a decision on pending the outcome of the 232.

Your next question comes from the line of Praneeth Satish with Wells Fargo.

Maybe just going back to Section 232. Obviously, there's a lot in play, and I recognize that. But we've heard and you mentioned the potential for waivers or quotas being allowed for certain domestic cell producers that could exempt them from some of these policy changes. I guess I'm just curious, conceptually, from your perspective, if some of these waivers are granted, do you think that could mute some of the price upside from Section 232? Or do you still see a constructive supply/demand set up? Just trying to think conceptually, how you think about that.

I mean, obviously, any modifications versus a 100% restriction will create some potential dilutive impact to the strategic incentive to 232. It also depends on if there is a waiver or some type of quota of some type. I mean, how big is it? Does it scale down over time? I mean, is it something that as it looked at it initially? And then that will walk down to maybe complete elimination of it.

So it's hard to give you a great insight to the impact. Clearly, we're not -- we're advocating to try to minimize any of those impacts and as well as they should only be a limited duration to the extent that they're enabled or allowed in all. We really want to create a domestic supply chain. And any type of workarounds that you get will disincentivize the investments that need to be made here in the U.S., right, to scale up those capabilities that -- and I think it's much easier for people to understand the policy environment certainty versus creating uncertainty by waivers of forwards and those types of things that they can create.

So we'll have to wait and see. We're firm in our position that we don't believe that they should be allowed, but we'll have to see how the final outcome is.

And there's some history here, too. If you look back at the Section 201 tariffs and the exemption was put in face-by-facial technology. It was clear that, that exemption effectively gutted that provision. So I think the administration has seen how those exemptions can effectively underline what they're trying to do. If there's a belief that the 232 provides a need around the national security interest, it doesn't make a lot of sense to have a carve-out or a quota piece associated with a national security interest provision.

Got it. That makes sense. And then if we say that Section 232 goes through, you get some kind of reasonable outcome, positive outcome. You kind of mentioned that there's 4 gigawatts. It sounds like 4 gigawatts plus of kind of pending deals for the second half. But do you get the sense that there's more demand sitting on the sidelines that's waiting for policy clarity? And once we get clarity, you could see that number move up significantly higher?

And then just a point of clarification. I guess, again, if Section 232 goes through, you get a good outcome. On the Southeast Asia capacity, would you bring that volume into the U.S. as finished products? Or would you -- would it come through as unfinished, and you would expand your U.S. finishing line?

So I guess, on the 232, and I'll let Alex take the other question around how we think through Southeast Asia and whether it comes in has finished or partially finished in order to expand capacity for finishing here in the U.S., I'll let Alex take that one.

The -- there are clearly are customers that are sitting on the sidelines. There is absolutely no doubt about that. And even some of these, that will -- even some of the stuff subject to CP is somewhat tethered to posting a security. So 1 of the challenges that, especially as you get longer dated in terms of contracting some of this volume -- and we are really trying to enforce having cash and liquid security against new bookings. That's been a priority of ours. In some cases, some of the counterparties can't post to the required security now, but working towards having that available, and makes sense that the security as opposed have been and kind of closes out of some of the CPs. So that's a piece of it.

But there's clearly people sitting on the sidelines, waiting to see what happens. We have a couple of counterparties that are they're hedging their way. And they know that the risk is the ASPs may go up. But at this point in time, they're trying to wait and see how it plays out. And again, kind of a conversation last time are the quotas or not and what are the options they have and so forth. So that's all in is in the mix right now. And as we've always said, the best thing for this industry is we just have clarity and certainty. 232, we just really need a decision on that because we can all [indiscernible].

As it relates to what we could do with the Southeast Asia facilities, we could bring fully finished product in subject of demand and pricing in the U.S. It's not only a function of where the 232 sits, it's also a function of where other tariff provisions sits. So right now, we have a Section 301 that's just gone into effect replacing the Section 122 tariffs that were in effect to the first half of this year. Those relate to forced labor. There is still risk around 301 related to excess capacity. So that investigation is ongoing. Pending the outcome of that obviously will determine what the total tariff impact could be then to product coming in from Malaysia, Vietnam.

We could bring some of it in as semi-finished WIP share product and finish it in our existing U.S. facilities. There's a limited amount, probably in the couple of hundred megawatt range of incremental capacity in our finishing lines across existing fleet in Ohio. So we could do a little bit of that. But it's not effective to run Malaysia at lower throughput as you're seeing with the unutilization costs we're having this year. So really, what we're looking for is an ability to run that factory at close to full capacity.

So then either it's selling fully finished international product, subject to where tariffs end up, or there is the potential to build another finishing line in the U.S. that's subject again to finding available site with power and the time it would take to build that out. So I think that's less likely, but it is still an option.

Your next question comes from the line of Julien Dumoulin-Smith with Jefferies LLC.

Quickly, actually to follow up on that last line of thinking on bookings. How to think about the safe harbor having played into the latest quarter here? Obviously, July 4 being a relevant threshold. And also, again, that being a leading indicator for future sales into the later part of the decade, how are you thinking about that? Obviously, that's a big part of your open book. What are you thinking in terms of having safe harbor go over your initial customer conversations?

And then as a follow-up, on what you were just alluding to there. Can you elaborate a little bit more around the permutations and the time line for that remaining piece in Southeast Asia? I know it's a little bit of just an extension of the logic you were just delineating there. But can you expand a little bit on the time line? It sounds like it's not that far off that you make a decision. Let me put it more bluntly.

Maybe I'll just take that one. On the Southeast Asia, we're really waiting for the outcome of the 232. We would expect to evaluate that and have a view shortly thereafter. It doesn't necessarily mean that we will have an immediate action plan that relates to, say, a shutdown or a full capacity. But once we have a sense of where the policy is, it allow us to evaluate it. It will take a little bit of time, though, we want to make sure whatever policy comes through, we understand it, and our customers also have a chance to evaluate it, and we can have discussions around whether there's a view of long-term offtake potential from those facilities.

Yes. And then on the -- I just want to make sure a couple of things. The bookings that we reporting most of the bookings that we recorded 1.9 of U.S. volume, I think almost all of that was outside of the quarter close. So most of that happened in July, which would also been outside of the safe harbor date.

And most -- everyone has safe harbor with transformers. There's really no safe harbor. I know there was a -- I don't know, it was like maybe 10 days left in the quarter where there was a ruling that was made that the decision that came out in August of the prior year, where it said that you eliminated the ability to use modular 5% CapEx rule to safe harbor. There was a ruling, but 1 of the courts that came out, I think, I don't know, somewhere like June 20 at that was hardly any time less than the quarter. In that theory you could use, assuming that, that wasn't challenged that you could use modules to potentially safe harbor projects. But I mean, that was really not an opportunity. It just happened way too late. And most of that already safe harbored with the inverters -- or transformers, excuse me, anyways.

But as you go forward, it is an important component, especially for anything that was safe harbored the first half of this year with the ability to COD them out in 2030, there are stricter requirements from a FEOC standpoint at the project level that has to be met that I think positions us well to serve that demand as you get out into '29 and '30 for when those products most likely could be commissioned.

Plus, the other thing I would say is we are seeing -- there's a lot of kind of rigid interpretations a little bit. And there are some people that are interpreting that even if something with safe harbor, let's say, in the second half of '25, that if you do anything with a change order or assume you move something from an MSA to [ APAPO ] or [ Tilia first store ], excuse me, is actually generated, you have to always be mindful of is there a restriction that you could have to comply with from a [ promotor an entity ] perspective.

So there's a lot of like very conservative, which is right so people want to be air tight and not taking any risk to jeopardize their either ITC or PTC. And I think there's a view towards maybe being overly conservative, advice are getting from tax counsel and others. And I think that's -- if I was in their situation, I clearly would do that as well. I don't want to put anything at risk.

So that safe harbor and those requirements under 48E as it relates to FEOC's restrictions or requirements I think will continue to play well for us as we look to book out through the end of this decade.

Your next question comes from the line of Philip Shen with ROTH Capital Partners.

Just wanted to follow up on the 232, specifically on timing. We've been thinking it's August, but we've seen a bunch of delays. The issue is if it slips past August, then we go into September and then get closer to the midterm. Then there's a chance that decision could push on that. Are they still [ honest ], but I wanted to [indiscernible] standpoint.

Phil, we're really having a hard time. You're breaking up.

Is it better?

Again, because it was really hard to get that.

[indiscernible] front and center. And so what's your view, based on the folks that you guys are in touch with, that this should be August? Or do you think there's a greater probability that this could slip into the fall or even beyond?

So I think I got the question. Look, we share -- look, there's -- I know there's a lot that's in the mix and what the administration is trying to evaluate when this is implemented, and we also want to make sure they do -- and what is implemented is achieves the strategic intent and the spirit of what it was set out to do.

So we are patient. We continue to be engaged. We are anxious as well as you are and others. And as I indicated, the industry really needs the certainty of understanding. I can't give you any level of conviction maybe more than what you have right now. We are still getting signaled that decisions will be made. There are meetings that are being had that would indicate they're close to making a decision.

But we also want to make sure that it is done right. And so to give you some sense of my level of confidence in August or whether it waits until September, I can't really give you a strong view on that. I can just tell you we want this to be implemented with the achieving strategic intended spirit of what it was set out to do. And that's the most important thing, and we're going to continue to be engaged with the administration to ensure that, that happens.

Your next question comes from the line of Colin Rusch with Oppenheimer & Co.

Guys, are there opportunities for you to reduce input costs on the U.S. manufacturing? And can you talk a little bit about the supply chain and how that's evolving? I know you had some discussions with [ glass bankers ] around capacity expansion and the capital needs that they have, but just curious about how you might look at that trend on a multiyear basis?

Yes. Colin, I mean, it's challenging. We're still in this -- especially in the U.S., as you see more reshoring, pressure on commodities, the data center is being built out. I mean, things, obviously, as you would expect, steel, aluminum, a little bit of copper. We don't use silver, but obviously, our competitors do. I mean, there's just a lot of pressure. Those -- but you can look at fuel costs, you can look at what's happened in the Middle East, and I see that as more transitory in nature. And they once that's resolved, and I think we'll see much more competitive fuel prices and what have you.

The electricity prices and some of the locations in which we operate, and we're dealing with some of those same adverse impacts that others are. So we're in a pretty challenging rising commodity cost environment. Now are we able to do things like drive more throughput through our operations? Absolutely, we're focusing on continuing to do that. Are we finding ways to create further automation and capabilities that can reduce labor costs? So there's levers that we're focused on. There's some redesign of the product that we're looking at and trying to take cost out of the back rails of the frame.

We continue to look at glass and thickness and other things that we could do from that standpoint. But it's a pretty challenging environment from a commodity cost standpoint. And our ability to get a lot of both that I think is probably 1 of the most challenging times that we've been in. Now I will say that when you look at it on the cost per watt, not necessarily our cost per module, the great thing about CuRe is that we have the opportunity to drive the efficiency up. So as we drive the efficiency up as we go from kind of where we are right now and add another 10, 15, 20, 30 watts, that will help the CPW numbers, right, cost per watt. Which is important, right? You didn't need to drive that number down. And then the ASP, the value uplift because of the energy attributes and the higher efficiency of CuRe that, that drives to an entitlement for higher ASPs and the like. So that's what we're focused on, and we're never going to give up on the input cost. We got to do the best we can to get cost out, but it is a pretty challenging environment right now.

I'd also say there's the potential to use the balance sheet to work with suppliers who are looking at expansion or needing funding. This an option there we could try and leverage our position of financial strength to get forward pricing that makes more sense that has to be done in the right risk premium risk profile.

And then the other thing I'd say is outside of just billed material costs. Obviously, we're having a challenging time around period costs going from cost per watt produced over the cost per watt what sold. So again, we're seeing freight challenges as it relates to cost of trucking. And I think I mentioned in the prepared remarks that we're seeing costs now to deliver product from Perrysburg over to the West Coast of the U.S., they're equivalent of delivering product from Asia to West Coast of the U.S. So we continue to look how we can optimize our domestic transport routes, freight, and try and ultimate between factories so that we can reduce those costs to grow as possible.

Our final question comes from the line of Corinne Blanchard with Deutsche Bank.

I actually wanted to come back on the last question regarding M&A, and I think you just added a little bit to it. But can you explain a little bit, what you -- what are you targeting with the current balance sheet that you have? And can't help feel like you were mentioning that you could use M&A to maybe help manage the input cost. But where else do you see maybe an option or a possibility for First Solar?

So when we talk about uses of cash, M&A is something that's been on the list for us for a long time. Generally, we focus more on the working capital reserve piece and then growing capacity and replicating technology. That's where the company has been, if you look over the last decade or so.

We've also put more money into R&D. And I think when you think about M&A, the obvious area for us to expand into would be do we spend more on and technology adjacent things, which could either be companies, it could be buying teams, it could be buying intellectual property, anything that could accelerate the technology transition we see going forward as we invest a lot into potential perovskite development. So I think there's options there.

We're also taking a look at things that are adjacent to the technology, but we want to do it with a disciplined focus around where do we see opportunities where we have a skill set that we can bring. So it's something where we look at our strength in high-volume thin film manufacturing and very high throughput efficiency. How can we leverage that set of skills and take it into an adjacent product, but also look at the overall market environment we're playing in.

We compete in a challenging industry where the vast majority of our competitors are Chinese and tend to play by a different set of rules. As we think about how we could move into adjacent areas across M&A, I want to evaluate what does the competitive landscape look like? What is the market that we will be accessing look like? What does the policy environment look like?

So we are starting to look through that. Clearly, given our position in the industry, a lot of stuff comes across our desk and has done over the last 10 years or so. We haven't done a lot on the M&A side. So we are more willing to do that. We're more open to it, but we want to make sure we do it with a disciplined focus.

We have reached the end of the Q&A session. This concludes today's call. Thank you for attending. You may now disconnect.

Hello, everyone, and welcome to First Solar's 2026 Annual Meeting of Stockholders. [Operator Instructions] As a reminder, place call is being recorded. I would now like to turn the call over to Mark Widmar, Chief Executive Officer of First Solar, Inc. Mr. Widmar, you may begin.

Thank you. I'd like to welcome you to the 2026 Annual Meeting of Stockholders of First Solar, Inc. Today's virtual meeting is a live webcast. We believe in engaging with our stockholders and hope that this virtual meeting will maximize the participation of stockholders regardless of their location.

Thank you for taking the time to participate today. On behalf of the Board, I also wish to express thanks to those stockholders who have submitted their proxies in advance. Participating with me today are Alex Bradley, Chief Financial Officer; Jason Dymbort, General Counsel and Secretary; and John Russo, partner, PricewaterhouseCoopers. We're also joined by the following directors: Lisa Kro, Curtis Morgan, William Post, Venkata Renduchintala and Norman Wright. I call your attention to the rules of conduct for this meeting, which are available in the meeting center.

We ask participants to abide by these rules so that we can conduct an orderly meeting. You can also find links to the annual report and the proxy statement in the meeting center by clicking on the document folder at the top of the screen.

Only registered stockholders are entitled to participate in the business of the meeting. The business to be conducted at this meeting is as follows: to elect 10 members of the Board of Directors to hold office until the next Annual Meeting of Stockholders or until their respective successors have been elected and qualified; to ratify the appointment of PricewaterhouseCoopers LLP as First Solar, Inc.'s independent registered public accounting firm for the year-end December 31, 2026.

To approve an advisory resolution on the compensation of First Solar's named executive officers and to vote on a stockholder proposal to improve shareholders' ability to call for a special shareholder meeting. While we'll be soliciting your questions throughout the meeting, feel free to post a question at any time. You may do so by clicking on the Q&A button at the top of the screen.

In accordance with the bylaws as appointed by the Chairman of the company's Board of Directors, I will act as Chair of this meeting. In accordance with the company's bylaws, Jason Dymbort, the General Counsel and Secretary of the company, will act as Secretary of the meeting. Christal Goldman of Computershare Trust Company, N.A., the transfer agent for the company, has been appointed the Inspector of Election for this meeting.

She has taken her oath of office, which will be filed with the records for this meeting. I'll now ask Mr. Dymbort give the Secretary's report on the qualification of this meeting to proceed.

Thank you, Mark. The notice calling this meeting was mailed on April 2, 2026, to stockholders of record as of the close of business on March 19, 2026. And the company has received an affidavit of mailing from the company's transfer agent confirming its mailing. The affidavit of mailing is available for inspection by any stockholder and shall be retained with the records of this meeting.

Since April 2, 2026, the proxy statement, the form of proxy and the annual report have been available to stockholders on the company's website in accordance with the Securities and Exchange Commission's notice and access rules.

In addition, the proxies and the certified list of stockholders are in the custody of the Inspector of Election. For more than 10 days prior to this meeting, the stockholders list was available for inspection by stockholders during normal business hours at our corporate headquarters located at 4300 East Camelback Road Suite 220 in Phoenix, Arizona.

The company's audited financial statements for the year ended December 31, 2025, are also provided on the company's website. Pursuant to the company's bylaws, the holders of a majority in total voting power of the outstanding capital stock of the company entitled to vote at a meeting of the stockholders present in person or represented by proxy constitute a quorum for the company's annual meeting.

Based on the inspector of election's preliminary report on attendance and proxy share count, I have been advised that a quorum is present at this meeting. When delivered, a copy of the Inspector of Election's report on attendance shall be retained with the records of this meeting.

As notice of this meeting has been given in accordance with applicable law and there is a quorum present, all legal requirements for holding this meeting have been satisfied. Following the formal portion of this meeting, Mark and Alex will present an overview of the company's business and address questions that are submitted.

Please note, this presentation and the ensuing question-and-answer session may include forward-looking statements that involve risks and uncertainties that could cause actual results to differ materially from management's current expectations. We encourage you to view the safe harbor statements contained in today's presentation for a more complete description.

Thank you, Jason. This meeting is lawfully convened and ready to transact business. The proxy statement made available to the stockholders in connection with the notice of this meeting sets forth the matters to be submitted for a vote of the stockholders.

At this time, if you have not voted or wish to change your vote, you may do so by clicking the link provided online.

[Voting]

Any stockholder who has already voted and does not want to change his or her vote need not take any further action. We now proceed to the matters to be voted on. The first item to be voted on in this election of 10 directors to hold the office until the next Annual Meeting of Stockholders or until their respective successors have been elected and qualified.

The proxy statement lists the director nominees, all of whom except Curtis Morgan, are currently serving as directors of the company. Each nominee has consented to act as a director for the ensuing year. The director nominees are set forth in the proxy statement are Michael Ahearn, Anita George, Lisa Kro, Curtis Morgan, William Post, Murthy Renduchintala, Paul Stebbins, Michael Sweeney, Mark Widmar and Norman Wright. Are there any comments or questions concerning the election of the directors?

Okay. The next item of business is the ratification of the appointment of PricewaterhouseCoopers as First Solar's independent registered public accounting firm for the year-end December 31, 2026. At this time, John Russo, a partner of PwC, is in attendance and available to answer any appropriate questions and to make a statement if he wishes. John, do you wish to make a statement?

No, thank you. I will pause now to allow for any comments or questions concerning the ratification of the appointment of PricewaterhouseCoopers as the company's independent registered public accounting firm.

Okay. The next item of business is the advisory vote to approve the compensation of First Solar's named executive officers. Are there any comments or questions concerning the advisory vote to approve the compensation of First Solar's named executive officers?

Okay. The next item of business is a stockholder proposal submitted by Mr. John Chevedden, requesting that the stockholders approve a proposal to improve shareholders' ability to call for a special shareholder meeting. At this time, [ Ms. Cam Franklin ] on behalf of Mr. Chevedden is in attendance via telephone, and I ask that she present the proposal. Operator, please enable Mrs. Franklin's telephone line.

Proposal 4, improve shareholder ability to call for a special shareholder meeting sponsored by John Chevedden. Shareholders ask the Board of Directors to take the steps necessary to amend the governing documents to give the owners of a combined 10% of the outstanding common stock the power to call a special shareholder meeting. Such a special shareholder meeting can be an easy to conduct online shareholder meeting.

There shall be no poison pill discriminatory rule to require ownership of shares for a specific period of time in order for shares to participate in calling for a special shareholder meeting. This proposal will give First Solar shareholders a genuine right to call for a special shareholder meeting to replace the current placebo right for shareholders to call for a special shareholder meeting that is more like an insurance policy that will prevent a special shareholder meeting from ever taking place.

The current First Solar so-called right to call for a special share meeting has 2 factors, each of which alone would likely prevent a special shareholder meeting from being held. The first factor is requiring the backing of 25% of shares to call for a special shareholder meeting instead of the 10% figure in this proposal.

The second factor is the First Solar disqualification of a substantial block of shares from being any part of the 25% requirement. Shareholders at more than 100 companies have voted on this proposal topic and not one of the 100 companies has ever given one example of a special shareholder meeting ever actually taking place at a company that had just 1 of these 2 factors in its bylaws for a shareholder right to call for a special shareholder meeting.

Please vote for an attainable right for shareholders to call a special shareholder meeting. Please vote yes, Proposal #4. Thank you very much.

Thank you, Ms. Franklin. Are there any comments or questions regarding the shareholders' proposal? Okay. I hereby declare the polls closed. The formal portion of this meeting is hereby concluded. The Inspector of Election will now tabulate the votes, during which time, Alex and I will present an overview of the company's business and answer any questions you may have.

We will then report the results of the voting before the close of the meeting. Beginning on Slide 6, our 4 growth pillars, a balanced business model and optimized product, a repeatable, scalable manufacturing format and continued investment in differentiation are how we deliver long-term shareholder value while navigating policy and market volatility.

On the topic of technology, turning to Slide 7. Our strategy remains anchored in a simple premise. Customers ultimately buy lifetime energy, not just nameplate efficiency. And our road map is designed to optimize the balance of efficiency energy yield and cost while leveraging our industry-leading thin film expertise.

Our technology strategy continues to be primarily concentrated on 2 core thin film focus pillars. First, we are executing a disciplined phase-gate introduction of CuRe, responsibly bringing this new technology to market with proven laboratory results and expanding field validation.

Consistent with our prior outlook, we are pleased to report that the CuRe launch is complete in Perrysburg, and our first Series 6 line is ramping consistent with expectation. CuRe is scheduled to be replicated across Series 6 and 7 fleet through the first half of 2028, which achieved supports the potential realization of up to $0.6 billion of additional revenue from technology adjusters in the backlog, with the majority anticipated in 2027 and 2028.

Our second pillar, Perovskite, is a key part of our effort to develop next-generation thin film semiconductors that can be deployed at commercial scale in both our traditional utility scale market while potentially expanding our addressable market segments.

To date, we have achieved reliability results that we believe are comparable to best-in-class R&D efforts while continuing to advance efficiency and stability, 2 of the industry's key hurdles for scalable Perovskite technology. A major enabler of these efforts is our dedicated Perovskite development line in Ohio. On Slide 8, responsible solar remains a core differentiator.

Our vertically integrated resource-efficient manufacturing delivers a higher energy result -- excuse me, return on energy invested than crystalline silicon alternatives. We employed thousands of Americans across 9 states and are proud to be the only solar manufacturer with more than 20 years of experience operating a global high-value PC recycling facility.

Our industry-leading closed-loop process recovers semiconductor materials for use in new modules and returns high-quality glass, rubber and aluminum to domestic supply chains. I'll now hand the call over to Alex to discuss our 2025 financial performance.

Thanks, Mark. Turning to Slide 10. 2025 was a record year on multiple fronts. Commercially, we sold record 17.5 gigawatts. Financially, we delivered record net sales of $5.2 billion and full year diluted EPS of $14.21, ending the year with a $3.4 billion net cash balance.

In manufacturing side, we produced 16.1 gigawatts, commissioned and initiated commercial production at our Louisiana facility and our South Carolina finishing line remains on schedule for Q4 2026. We also saw strong performance in the first quarter of 2026 with record first quarter revenue of $1 billion, record volumes sold of 3.8 gigawatts, including record sales in India, meaningful margin expansion and record Q1 diluted EPS of $3.22.

With that, Mark, I'll turn the call back to you.

Okay. The floor is now open for any questions. You may submit questions online by clicking on the Q&A button at the top of the screen. Are there any questions that any of the stockholders would like to ask at this time?

Mark, there have nothing -- any questions submitted by stockholders.

Okay. If there are no questions or comments, I will ask the Secretary of the meeting to read the preliminary voting results.

On the proposal regarding election of directors, each nominee recommended by the Board of Directors has received a majority of the votes cast with respect to such candidate's election. Therefore, each of the 10 nominees has been duly elected to hold office until the next Annual Meeting of Stockholders or until their respective successors have been elected and qualified.

On the proposal concerning the ratification of the appointment of First Solar's independent registered public accounting firm for the year ending December 31, 2026, 79,960,750 shares or approximately 86.8% of the shares represented and voting on this proposal have voted for ratification of the appointment of PricewaterhouseCoopers LLP as the independent registered public accounting firm for the company for 2026.

On the proposal concerning the advisory vote to approve the compensation of First Solar's named executive officers, 69,188,428 shares or approximately 86.81% of the shares represented and voting on this proposal have voted on an advisory basis to approve the compensation of First Solar's named executive officers.

On the stockholder proposal to improve shareholder ability to call for a special meeting -- sorry, 34,307,259 shares or approximately 43.05% of the shares represented in voting on this proposal have voted to approve the stockholder proposal to improve shareholder ability to call for a special shareholder meeting. Accordingly, the proposal has failed to pass.

Mr. Chairman, the voting results announced today are preliminary and subject to final tabulation by the inspector of election.

The final results of the stockholder vote reflecting all proxies received through the close of the meeting and any votes cast during this meeting will be included in the final report of the Inspector of Election and will be published on a Form 8-K to be filed with the Securities and Exchange Commission and will be available upon request.

Thank you, Jason. This meeting is now concluded. I want to thank you all for attending today's meeting and for the support you have shown First Solar.

Thank you, everyone. This concludes the meeting. You may now disconnect.

Good afternoon, and welcome to First Solar's First Quarter 2026 Earnings Conference Call. This call is being webcast live on the Investors section of First Solar's website at investor.firstsolar.com. [Operator Instructions] Please note that today's call is being recorded.

I would now like to turn the conference over to First Solar Investor Relations.

Good afternoon, and thank you for joining us. We're joined today by Mark Widmar, our Chief Executive Officer; and Alex Bradley, our Chief Financial Officer. Mark will provide an overview of our first quarter performance and an update on technology, manufacturing and market conditions. Alex will then cover our bookings, financials and our 2026 outlook. After our prepared remarks, we'll overline for questions. Today's discussion contains forward-looking statements. Actual results may differ materially due to risks and uncertainties as described in our earnings press release, other SEC filings and the earnings materials available at investor.firsolar.com. We undertake no obligation to update these statements due to new information or future events. We will also reference certain non-GAAP financial measures. Reconciliations to the most directly comparable GAAP measure are in our earnings press release and presentation. This non-GAAP financial information is not intended to be considered in isolation or as a substitute for financial information presented in accordance with U.S. GAAP.

With that, I'll turn it over to Mark.

Thank you, and good afternoon. Beginning on Slide 4. We delivered a strong start to 2026 with record first quarter revenue, record sales in India, meaningful margin expansion and adjusted EBITDA above the top end of our first quarter preview rate. Since our last earnings call on February 24, we secured gross bookings of 1.9 gigawatts, excluding domestic India volume, we booked 1.4 gigawatts into our key U.S. utility scale market at an ASP of approximately $0.35 per watt, inclusive of applicable adjusters.

Turning to Slide 5. Our technology strategy is anchored in the premise that customers value not just nameplate efficiency, but lifetime energy production as well. CuRe is central to that strategy. Extensive testing data has validated our expected bifaciality, advantaged temperature coefficient and degradation profile. With CuRe anticipated to deliver up to 8% more lifetime specific energy yield than crystalline silicon Topcon. I'm pleased to report that CuRe launch is complete in Perrysburg and the first Series 6 line is ramping consistent with expectations. CuRe is scheduled to be replicated across the Series 6 and 7 fleet through the first half of 2028 which, if achieved, supports the potentialization of up to $0.6 billion of additional revenue from technology adjusters in the backlog with the majority anticipated in 2027 and 2028.

Turning to Slide 6. We produced 4.3 gigawatts of modules in the quarter with approximately 3 gigawatts from our U.S. facilities and 1.3 gigawatts from our international fleet. Our U.S. facilities operated at approximately 96% utilization. South Carolina finishing facility is on track for production start in the second half of 2026, with equipment installation beginning this quarter. Upon completion, the facility is expected to provide finishing capacity for Series 6 modules initiated at our international factories and optimized for freight tariff and domestic content outcomes while benefiting from Section 45X module assembly tax credits.

Our international facilities in Malaysia and Vietnam continue to operate at a significantly reduced utilization consistent with current trade dynamics and lower ASP expectations for internationally produced modules. Turning to Slide 7. Our competitive position in the United States and India continues to strengthen, underpinned by differentiated technology, a domestic manufacturing footprint and bill of material, and independents from Chinese crystalline silicon supply chains. In the United States, headwinds for crystalline silicon continue to build in our view, including trade remedy enforcement, indications of restricted FEOC regulations and the intellectual property litigation actions we have discussed on recent calls.

Our IP specifically in March, the U.S. International Trade Commission instituted our Section 337 investigation with respondents representing a significant share of top 10 modules currently imported into United States. We expect an initial determination within approximately 11 months and final decision within 15 months. In India, our presence reflects the same strategic logic that underpins our U.S. manufacturing investment, energy security and supply chain independence. The policy framework, including the existing approved list of models and manufacturers or ALMM and the anticipated implementation of the ALMM at a cell level as well as domestic content requirements currently favors manufacturers such as First Solar.

Near-term demand is supported by both utility scale and distributed solar applications, including agricultural land developments where our technologies energy yield in hot human conditions is a meaningful differentiator. Overall, our differentiated technology, our domestic manufacturing footprint, and our independence from Chinese supply chains are attributes that are increasingly valued by our customers and we remain well positioned to deliver on our 2026 commitments.

I'll now turn the call over to Alex to discuss our bookings, financial results and outlook.

Thanks, Mark. Beginning on Slide 8. As of March 31, 2026, our contracted backlog was 47.9 gigawatts and an aggregate transaction price of $14.4 billion, exclusive of technology adjusters with deliveries through 2030. During the first quarter, we sold approximately 3.8 gigawatts, reported gross bookings of approximately 1.7 gigawatts and recorded debookings of 0.1 gigawatts. In India, our guidance assumes that production is largely sold domestically in a book-and-bill market at near full capacity. In the first quarter, we sold approximately 1 gigawatt in country at an average selling price of approximately $0.20 per watt. .

In the United States, our domestic brand is substantially committed through 2028 under existing contracts, resulting in relative pricing clarity through this period. We continue to take a highly selective approach to incremental U.S. bookings as we await the outcomes from current policy and regulatory matters and particular, the pending 232 polysilicon derivatives tariff decision and proposed FEOC rule-making. During the first quarter, our gross bookings of 0.9 gigawatts were at an average selling price of approximately $0.34 per watt inclusive of applicable adjusters.

With respect to our international fleet, demand for Series 6 modules produced and end in Malaysia and Vietnam remains constrained, which is reflected in the reduced production Mark mentioned earlier. Turning to Slide 9. Net sales of $1 billion were a record first quarter for the company and grew 24% year-over-year. This is [indiscernible] by a 31% increase in volume, partially offset by lower average sales price reflecting a higher proportion of India deliveries. Our gross margin in the first quarter was 47% and expanded approximately 6 percentage points as compared to the first quarter of 2025. The drivers were primarily a higher volume of modules qualifying for Section 45 tax benefits and significantly lower sales rate costs, including lower detention and demurrage.

On a per watt basis, sales rate costs for approximately half of first quarter costs last year at approximately $0.017 per watt. Furthermore, as part of our planned rationalized warehouse costs to approximately $100 million by 2027, we delivered a $22 million sequential reduction in warehouse costs from Q4 2025. On balance, these savings were partially offset by a lower average sales price due to a higher mix of India sales and increase in tariff costs year-over-year. Operating expenses for the quarter were $141 million, including R&D of $67 million, up $15 million year-over-year, primarily reflecting [indiscernible] development and ongoing CuRe launch.

Adjusted EBITDA was $520 million, above the high end of our first quarter preview range of $400 million to $500 million. Adjusted EBITDA margin was 50%. Net income was $347 million, up 65% year-over-year with diluted EPS of $3.22, Moving to Slide 10. We ended the quarter with $2.4 billion of cash, cash equivalents, restricted cash and marketable securities and a net cash position of $2 billion, at the high end of our targeted resilient net cash range of approximately $1.5 billion to $2 billion. Operating cash outflows of $215 million reflected normal first quarter working capital dynamics, a meaningful decrease from outflow of $608 million in the first quarter of 2025. Capital expenditures were $119 million, primarily for our South Carolina finishing facility. We also completed a $45 million scheduled principal payment on our India DFC loan.

Turning to Slide 11. Our full year 2026 guidance remains unchanged. For the second quarter, we expect volumes sold between 3.4 and 4 gigawatts and adjusted EBITDA of $400 million to $500 million. In summary, our first quarter performance and reaffirm guidance for the full year reflect the strength of our strategy of reshoring and scaling domestic manufacturing, progressing our technology road map, enforcing our intellectual properties and maintaining a selective approach to new bookings in light of key pending trade and policy determinations.

And with that, we conclude our prepared remarks and open the call for questions. Operator?

[Operator Instructions] Your first question comes from the line of Brian Lee from Goldman Sachs.

Just had 2 here. First, on the module gross margins, I think it's around 7% in Q1 when we adjust out the 45X even with the high India mix. So curious, the guide for Q2 implies margins are flattish. Why not more improvement given some of the sequential improvement in freight costs, warehousing, et cetera. And then what kind of tailwind maybe help into 3Q and 4Q for the margins beyond just volume growth? And then on the ASPs, maybe I'll just squeeze in the second question here. This could be nitpicking, but on Slide 8, you show 900 megawatts of U.S. bookings, $0.34 in Q1 and you mentioned 1.4 gigawatts since the last quarter at $0.35. So it maybe implies recent bookings. March and April are higher at $0.36, $0.37 per watt. Anything to read into that? What kind of customers engagements and discussions that you're having here more recently ahead of more policy certainty.

I'll do the ASP first, Brian. And then ask Alex to go through the gross margin. ASPs, yes. So the $1.4 billion is call the call. So that's a little bit of maybe a clarification there. So -- which would include half and half. So half of that happened before the quarter end, all that happened after the Feb call, about half of it happened before the quarter end, the other have happened after quarter end and the average ASP for call-to-call of that $1.4 billion was $0.35. One of the notes to include in there is that there's about another 700 or so that's an option. We're seeing a lot of M&A activity. We're seeing a lot of -- what's great about First Solar. We've had very strong strategic partnerships with very confident obviously, well-capitalized partners, and they're actually seeing a lot of development acquisition opportunities, actually talking with the team about another deal right now that 1 of our customers is actually in the process of acquiring is looking for incremental volume to support that acquisition.

So what happened was we booked a deal for about 700 megawatts. Our customer is actually now in the process of looking to acquire another development asset, of which then they would exercise that option, which would have to be exercised here over the next several quarters upon completion of that acquisition. So what I would just say, there's still a lot of momentum, a lot of activity going on, Brian, from a market standpoint. We're being very disciplined as we have in the prior quarters with how we're engaging the market and how we're seeing through pricing. So still good momentum, discipline on our part trying to realize good ASPs, which I think we did here in the last, call it, few weeks since the last earnings call.

Brian, on the gross margin, so if you think about India on an aggregate sense per watt or dollar contribution basis, it's certainly lower than the U.S. If you look at where India's pricing today on a gross margin percent basis, it's not materially different to the U.S. So despite having a high India mix on the percentage gross margin basis, not a material impact. If you look at the full year guide of 7%, we haven't changed the guide, that still holds. That's right where we came in ex IRA benefit in Q1. as you think through going forward, next quarter, we're guiding to the same guide that we had in Q1. So it's relatively flat, which implies the second half is going to be stronger. Incremental volume should be beneficial to gross margin. There's a little bit of value on the fixed cost side. The other piece in terms of the back end of the year is that right now, our assumption on tariffs is that the Section 122 tariffs carried through for the 150 days from announcement takes us through to the July time frame.

After that, we're not modeling tariffs beyond that for finished goods coming in, right? There's still other tariff impact in there. Now that's the modeling assumption we have today given uncertainty around what could happen. As you're probably aware, the Trump administration has launched several 301 cases with the intent, we believe, to try and replace the 122 with that 301 later in the year. But then now we're not modeling that. So I think the guide has stayed where it is, the guide assumes our tariff replacement back end of the year, that would imply you might get some incremental gross margin. However, if we do see additional tariffs [indiscernible] to reflect that in the guide.

And I think maybe 1 of the things, too, just around operation quarter-on-quarter from a gross margin standpoint, we will be running -- we ran Malaysia, Vietnam at higher utilization rates in the first quarter than we anticipate running in the second quarter. So you're going to see more underutilization charges in the second quarter. So trying to get to your question, Brian, about sequential gross margin performance, we will see a little bit of headwind in the second quarter because of the lower utilization rates for Malaysia, Vietnam.

Your next question comes from the line of Julien Dumoulin-Smith from Jefferies.

This is [indiscernible] here for Julien. I guess 2 quick ones. Could you quantify the amount of adders, I guess, on that $0.34 or $0.35 ASP? Is it like 2 [indiscernible] that you discussed before? And then second, a real quick one. How do you think about the Southeast Asian capacity going forward?

I'll take that. I'll do the adders and Alex, you can talk to the Southeast Asia thought process where we are on that. So 1 of the things is now with the launch of CuRe as an example, we will -- you'll see it's kind of -- I want to make sure people understand is the product which we're going to price now going forward, given that the CuRe product is being launched is going to be the technology which we anticipate to deliver in this full entitlement. And if you look at the bookings that we reported, call at 1.4 gigawatts since the last earnings call. Half of that volume actually sits out in 2029. So it's encouraging, we're starting to see more momentum in the outer years.

And that's a CuRe product that we expect to deliver. And so when you look at the adders and that in some cases, the volumes that we're seeing right now, there are no adders because we're pricing the technology, therefore, you won't see an adder to that deal. And so when you look at the blended average of the adder -- the [indiscernible] of the adders are still, call it, $0.03 or so, but half of the volume that we booked did not have adders on it, the other half did. So the blended average of the adder is going to be call it, $0.01 or something a lot of those lines. But I think what we'd like to be able to do as we move forward, especially now with the launch of CuRe. We're pricing the technology. We're going to deliver the technology, we're going to price it.

All the energy attributes will be embedded in the base price. We're kind of still in that transitionary period. We're going to see some contracts that potentially have the combination of 2, the base being our current semiconductor, which we refer to as [indiscernible] puls adder, but for the windows in which we for sure are going to be delivering the CuRe product will just price it as the contract, and you won't necessarily continue to see the adders. I just try to make that clarification. So as we transition in that direction, you're probably going to see less volume with adders on and just full entitlement of the technology being priced.

As it relates to Malaysia, Vietnam, we talked on our previous call about maintaining an option around that capacity, and we're still doing that. If you look through what we're waiting. I think a large part of that is policy clarity around the 232, which really could spur potential demand around the fully finished international product. If you go back to the original guide, we had $115 million to $155 million of underutilization costs. That was a function both of Malaysia and underutilization running at very low capacity through the year as well as some of those costs associated with the new finishing line that we're bringing up. So right now, we're continuing to maintain that near-term option. We'll continue to evaluate that. Likely, the decision point we're waiting on is positive around the 232, which you expect most likely to come in Q2 of this year.

Your next question comes from the line of Christine Cho from Barclays.

I actually wanted to know if we should think that there's an impact from the Section 232 changes on steel and aluminum for the Southeast Asian imports, I was under the impression that aluminum might be less than 15% of the weight of a full module, but now that you're going to be just importing just the front of the module, like should we think about that? And then I think you also said on the last call that 5 gigawatts of the backlog was international modules. Was the plan just to mostly import that at the beginning of the year and it tails off at the end of the year? If you could just talk about the cadence of that as well.

So I got the [indiscernible] repeat your second question, I'll make sure I have it.

My second question, the 5 gigawatt of backlog was -- I thought that was the international portion. Was it not?

Yes, yes. No, that's about right. I think that's approximately number. So yes, so that was entering the year, we had gigawatts of 6 international backlog. And so from there, your question was?

Was the plan to mostly import that at the beginning of this year?

No, no. So okay, I got it now. Sorry, that's a piece I didn't have. Yes, that 5 gigawatt of backlog on S6 international was a multiyear backlog. Those shipments would go across '26, '27 and into '28. So that was a multiyear. So we will run that production to meet that demand of that portion of the backlog, call it a little bit more than 25%, around 25%. It would be for this year. The balance would sit in the outer years as it relates to that. On the 232, yes, aluminum is still included in the tariffs for 232. I think part of your question was as well, what happens with the semi-finished product that comes into the U.S. That will come in -- the glass will come in, obviously, without the aluminum frame. As of now, we will continue to import the aluminum frames into the U.S. And because they continue to be imported, they'll be subject to the applicable rates associated with the 232 for aluminum. So anyway, yes, 232 is still applicable for or aluminum based on the classification of the product we're bringing in and no real change to that just because we're bringing in a semi-finished product from Malaysia.

Okay. And then you guys have -- your -- you guys have been doing a lot of India volumes in this quarter and the full year assumes full utilization. I just saw that there was like in 1 of your disclosures about India's there's a proposal in India to increase the minimum efficiency of PV modules for manufacturers to be included in the ALL beginning in 2027. Could you talk through exactly what's going on there and what your options are to work around it? I realize that it's still just a proposal, but?

Yes. A couple of things. There's a lot of moving things that are going on in India, even including the requirements by 2028 to have a qualification of the wafer being domestically manufactured in India in order to qualify for any projects that actually connect to the grid, The federal, so that's moving as 1 of the moving pieces that will, I think, further enhance our opportunity to continue to support the India market given a vertically integrated model. The other 1 as it relates to the consideration for the efficiency threshold. We will be launching CuRe beginning of next year. in India. So our first Series 6 facility that we'll launch in India will be CuRe, which will give us an opportunity to improve the efficiency of the technology as well as continuing to enhance the energy attributes better bifaciality, better temperature coefficient, better long-term degradation rate.

And as we continue to work with MNRE, in particular, in other parts of the administration and continuing to inform them and educate them on the value of the attributes of energy, which is also what our customers pay for energy, not labeled efficiency, We continue to have very good and constructive dialogue in that regard. But I think the key enabler to make sure we manage through any potential revisions to those requirements will be the launch of our next-generation technology of CuRe in India.

Your next question comes from the line of Chris Dendrinos from RBC Capital Markets.

I just wanted to follow up on the earlier question from Jefferies in regards to the Southeast Asia offtake agreement. Can you maybe walk through a bit of the possible decision tree here, depending on, I guess, the tariff outcome or an offtake agreement and how you're thinking about that facility? And when would we have a potential decision as to what you all do longer term?

Yes. I mean right now, there's ample demand for the product at a certain price, but when you factor in the tariff implications of bringing that product in and then the risk allocation around that tariff, it's not necessarily the right risk profile for us to take today. Depending on an outcome of the 232, you could potentially see much higher pricing or risk tolerance from buyers whereby we will be able to more appropriately price that product. And more allocate the risk around changes in tariff policy, which will then enable us to be comfortable long-term contracting that product.

So a lot of it depends on availability of supply in the U.S. and where tariff and risk can be allocated. And so we think the 232 is most likely the main determinant of that. The outcomes there could be we continue to run that product at full capacity end to end and ship fully finished goods into the U.S. it could be that there's demand, we could add an incremental finishing line in the U.S. and finish that capacity here? Or it could be that neither occur and then we're into the potential shutdown of that capacity. So those are really the pieces we're looking at.

Yes. And 1 thing maybe just to help -- remind everyone as we showed in the last earnings call. So historically, we've had about 7 gigawatts of capacity between Malaysia and Vietnam, okay? Half of that is now going to come to the U.S. to support our South Carolina finishing line. So 3.5 of that that leaves the other 3.5. Now what we indicated in the last earnings call, about half of that 3.5 is largely no longer available because we're moving some of that back-end capacity to the U.S. to support our Purovsky pilot line. So Purovsky full-size Series 6 pilot line will be available in 2027. So we lose the back-end capacity. As a result of that, we're losing some of that throughput for the facility.

So when you really look at how much of the Malaysia facility of that 3.5 gigawatts that would be available, which will be fully manufactured, assembled modules to be shipped into the U.S., there's only about call it, 1.8 gigawatts, maybe closer to 2 of real capacity. So what we're talking about is from a full-size finishing module capacity perspective, there's slightly less than 2 gigawatts that's in play that will tether back to a 232 decision point. So I just want to put that in perspective. As you think through your analysis and assessment of Malaysia, Vietnam half of it's already going to be coming to the U.S. as a semi-finished product. Half of it has already been -- capacity has been reduced because we're moving the back-end tools to help support our [indiscernible] pilot line. And then now we're kind of still with about slightly less than 2 gigawatts in Malaysia, Vietnam, but it will be tethered back to whatever decisions made with 232.

Your next question comes from the line of Phil Shen from ROTH Capital Partners.

On the 232, I was wondering if you might be able to share a little bit more color on what the framework of that decision might be. I think we published earlier this week that there could be a minimum import price in the $0.38 per watt kind of level. And so does that resonate with you guys at all? And on your Slide #7, you talked about the timing being Q2. We've seen this push a bunch, right? It was supposed to be year-end '25, but then the shutdown happened and a bunch of other reasons have driven this a little bit later. And so as we are a month into Q2, what's the confidence level that it comes out in May or June? And then on the technology front, Mark, I think you just talked about a full-scale line of perovskite in 2027, which is really interesting. And so I was wondering if you could give us a little bit more color on that? What kind of costs are you seeing? And then is that -- what's the base. I believe it's CDT on the bottom cell and then perovskite on the cell. But just any other color in terms of efficiency, durability, et cetera?

Yes. So Phil, on the 232 and the framework, I mean, look, there's still a lot of moving pieces, right? What all I can say right now is that the engagements that we're having with the administration and the structure of which we are trying to propose, which again is not 1%. It's basically $0.01 per watt on the cell or $0.01 per watt on the module or could include a minimum import price to your point. the feedback we continue to get is very positive to looking at that as the best way to address the polysilicon and its associated dervaties. So it's a good positive feedback.

But Phil, as you know how these things play out, they continue to evolve. And so we have to stay as well connected as possible to continue to advocate, advocate and advocate for that type of position. But what I would say is, at least, is it's encouraging. As it relates to the timing, similar the feedback that we are getting is still a resolution by the end of this quarter, does it move into early Q1 potentially. But it's all dependent on other events that happen, right? There's always I think the intention is that they know they need to bring this to a conclusion. It's important and they need to communicate an outcome. But as you know, Phil, these things can always move.

So is there a level of certainty for the quarter, what we just represented in terms of the slide is the information, the best information we have and what our expectations are based on those communications around an outcome and communication around 232. On the technology side, yes, so we -- we are currently in a time line that would have us running a pilot line in 2027 for prostate. The cost piece of it, I'm not going to get into the specifics of that. But that pipeline, which is a 1 gigawatt pilot line, by definition, it's not going to be an HVM type of cost entitlement, right? Because in order to get cost out, you need to run high throughput and scale, right? That's like if you go back and if you -- if you look at when we went from our Series 6 factories to our Series 7 factories, they're -- we tripled the output effectively and we managed it with the same headcount. So no real dramatic change to headcount, and that helped drive cost out significantly.

And so 1 gigawatt, you're going to be suboptimized and cost is going to be higher, but for an initial product to get it into the market and to get field validation and customer feedback and all the other things that we need, we think it's appropriate to do that upon launch. And then as we continue to evaluate, we'll move into a scaling mode. But again, it's going to be a higher cost product. There's no doubt about it upon launch. As it relates to the construct of what the product is? And is it a single junction, is it tandem? I mean there's -- we're looking at 2 different paths. We're still evaluating what we believe is the right launch product. I think the most important thing to do to get something in the field is to validate the performance out of the perovskite. If you can do that and reduce some complexity of thinking through the additional challenges of integration of the tandem construct because you have 2 products that have different electrical properties, right?

Different temperature coefficient as an example, in different electrical properties, voltage current go on. to try to deal with the construct of matching those 2 different semiconductors and optimizing them and harmonize them at the module level. So they perform as effectively a system if you want to refer to it as such. There's a level of complexity around that, that I'm not sure it's worth the effort initially because what you want to do is get the product in the field, validate its performance, validate its diversion, validated performance in all forms of conditions, right, which would include open circuit, how does it perform in partial shading. So there's a lot you can learn.

And I think the most important thing you want to learn is at the perovskite level. We all know how -- whether the tandem construct would be a perovskite with a [indiscernible] modular with a perovskite silicon cell. Those you can learn and evolve, but what you really need to do is validate the durability and viability of the perovskite and they can perform in the field and be a bankable product and can hit the long-term performance objectives that is going to be healthy.

Your next question comes from the line of Vikram Bagri from Citibank.

My first question, Mark, probably for you, We understand the market is for contracting panels at $0.33 a watt or higher is not as deep customers are hesitant in pricing the Section 232 risk as of now. I was wondering like once it comes out based on your assumption sometime in [indiscernible]. How quickly can you move to book volumes, put it another way, how much demand is beating for Section 232 to come out? What's sort of like what should we see or look for in bookings after immediately after the Section 232 comes out?

Well, look, I think there is this some unknowns in there as well. I mean, once it gets communicated and the other is to what extent is -- what is the impact right? And that will be a key component of how much volume and how quickly. I mean what I will say is that we have some customers that are looking at multiple gigawatts of volume and they're waiting. And what we've also said to some of those customers that once this gets communicated, depending on what the outcome of that communication is it could have an impact on the current price that we're at least negotiating. So the risk we run is that it ends up being lower than anticipated, the risk they run is it ends up being higher than anticipated.

So I think there's quite a bit of demand there that should provide an opportunity for us to move through and to book that over a multi-month period of time. But it really depends on the outcome, and that's really what the -- where you got to bid ask right now. That's what it relates to is what do we expect that outcome. We're trying to create a price point, we think that's reflective of the midpoint and we'll see what comes out. If we're wrong, then we may see a little bit of ASP pressure. And if we're right, we may see a little bit of upside relative to that marker we have engaged the market with right now.

Your next question comes from the line of Colin Rusch from Oppenheimer.

It's a 2-part question. So first, as you move from Series 6 to 7 with CuRe, can you talk about the key technical elements that you guys are working on right now and things that we should be watching for, for success? And then can you talk about any sort of impact that you're seeing from the incremental capacity, domestic capacity that's coming online to some of your pricing negotiations?

Yes, Look, the -- from a tactical standpoint, Series 6 to Series 7 isn't really a significant technical challenge. What it is, though, is -- because it's just a form factor change for the most part, right? We are -- for the Series 7 launch -- so right now, we are for the front contact buffer, we actually are taking our existing TCO glass and then we're actually depositing the front contact buffer for S6. But for Series 7, our plan would be because we've been working with our glass suppliers to allow them just to do it within their facility. So we no longer we'll just get our standard TCL. We'll get the glass will also be included -- include the revised front contact buffer that we need for CuRe. So that will simplify the factory operations from that standpoint. But really, it's just the difference of the size of the tools.

So the tools that -- there are a couple of new tools that we've had to add, it's because of the BKM, best-known method and process is different for CuRe and our existing product. And so there are a couple of new tools. Again, those are first of the kind tools that have been designed and spec for a smaller form factor product in Series 6. So we have to now operate those tools, season those tools to get them to the performance level that we need for a different size form factors. So it's not as much of a technical challenge. It's just that there's a slightly different process that we're using because in Series 6 and at least the launch, we're doing the front contact buffer in Series 7, We won't be doing it and then the difference in size of the tools. is really what the real challenge is. So which -- we're working through that. We're validating all that right now. And we're going to go as fast as we can.

Once we get comfortable with a validation of a lot of that, we'll move forward and try to replicate as quickly as we can across the fleet. The new capacity, not -- so if you exclude our fully integrated capacity between Ohio, Alabama and Louisiana, the new facility in South Carolina is going to be semi-finished product. So what's nice about it is it's a Series 6 form factor. It does have domestic content. It doesn't have as much domestic content is a product that we manufacture in Ohio, but it creates a nice opportunity to blend the 2 together for our customers. They get value out of that as well. it does sort of create an opportunity to be a little bit more competitive on pricing.

It creates an opportunity to create more domestic content value to our customers and therefore, enabling a broader portfolio projects that can benefit from the domestic content bonus, which is extremely valuable to our customers, especially on the ITC side, which in some cases, we're talking at the project level, $0.15 or $0.20 or more of value to our customers to enable that bonus. And so it's a nice way to sort of enhance it further create value to our customers' portfolio by enabling them to see higher realization against domestic content bonus for ITC in particular.

Your next question comes from the line of Praneeth Satish from Wells Fargo.

With the IPA tariffs repealed and import tariffs on India produced modules down to 15%, just seeing how are you thinking about selling the India capacity into the U.S. versus selling it in the Indian market? Is that something that you're still considering? Maybe if we get a positive 232 outcome and then to the extent that you are, what's the lead time and retooling costs that would be required to enable that? And then just a quick housekeeping question on the 1.7 gigawatts of bookings this quarter. Are you able to break out roughly how much of that is from U.S. capacity versus international?

The -- on the India -- I mean, right now, there's a lot of demand in India. We just sold 1 gigawatt last quarter. And when you look at the actual gross margin on that product effectively the highest gross margin that we have, including the benefits of 45X. So the gross margin is on a percentage basis is obviously attractive. The -- and the changeover is always -- downtime -- even though we try to optimize to make it more efficient to change from a fixed sale product to a tracker product. It still is not efficient to do that. So we are right now looking at this from a lens, let's just keep running that product because of the demand that we're seeing, at least through the first half of the year. Q3, we see a little bit of softness generally in India, and then you'll see a stronger Q4.

So there maybe a little bit of volume in three, assuming the tariff environment stays where it is, the challenge with the 122 expire at the end of July. So we're going to be looking at whatever tariff environment would be enduring after that. I don't -- we have no idea what will happen with the 301s. And so it's kind of a watch card. We'll see what happens. We will be bringing some volume over, we're talking 10s to low 100 megawatts or so of whither product from India and bringing that into the U.S. and finishing here in the U.S. We're doing some of that in the first half of the year to help enable demand here in the U.S. market. But yes, it's something we continue to evaluate. I just think we don't want to -- until we have a better understanding of what the long-term tariff environment is going to be because, again, we don't know after the 150-day window for the 122 is what exactly a tariff environment will be in. I think for now, we're going to focus on a market that has very strong demand continue to support it from that standpoint.

Praneeth, your housekeeping question, the 1.7, $0.9 million was U.S. and $0.8 million was India bookings.

Your next question comes from the line of Maheep Mandloi from Mizuho.

Just maybe some housekeeping on the first half or the second half cadence, if I look at the EBITDA versus the [indiscernible] looks like the EBITDA is 36% in the first half, but 44% of the volumes. So just trying to understand, is that because of India shipping is higher in the first half? Or is it SG&A, which is kind of skewing the EBITDA in the first half or second half?

Yes, it's mostly driven by the India volumes. As Mark said, we had a strong Q1 for India. It will drop down in Q2, Q3 and potentially pick back up in Q4. But as of now, the guide assumes that imbalance, which is why you're seeing that.

Your next question comes from the line of Joseph Osha from Guggenheim Securities.

I'm still trying to understand the composition of this 3.5 gigawatts in South Carolina. Are you saying that we have for starters perovskite part of that? And then you're saying that absolutely no matter what happens, we're going to run 1 gigawatt and changes to it through S6 and the rest is optional. I'm trying to put together where this 3.5 comes from.

Yes. Thanks for that question, Joseph. That 3.5 in South Carolina is our CuRe or will be our CuRe product and Series 6 product has nothing to do with perovskite. So it's going to be a product that will be started, think about it just basically the substrate glass with the deposition on it with cell scribing and then ship to the U.S. to be finished, which will include the cover glass junction back -- junction box, the frame, [indiscernible] all the other stuff that's going to be added to it, right? So that's what that product is, right?

So it is a [indiscernible] product, okay? In addition to that, we will be launching a pilot line next year, which we communicated this as well on our last earnings call, we made the announcement that we acquired the IP for Oxford for perovskite which enhances the IP that we already have to manufacture that product here in the U.S., and we will be starting that with the pilot line that will have up to 1 gigawatt of capacity. That pilot line will actually be in our Perrysburg facility. So we have existing facility space in Perrysburg that we will be integrated and using for that pilot line. and we'll be leveraging some of the capabilities that we already have from a back-end processing and what have you in Perrysburg. So that's what that means, right? So it's a pilot line. We still think of it as development has a gigawatt of capacity. We will be manufacturing product when we deploy it into the market, getting test data validation with customers and those types of things. South Carolina will be a [indiscernible] product, started in Vietnam and Malaysia, finished in South Carolina.

And maybe just to the confusion there. So if you think about the original 7 gigawatts of capacity we had in Malaysia, Vietnam, 3.5 is still going to be used for the front end that the product was comes to be finished in South Carolina. Of the remaining 3.5 we're using some of those tools. Those tools will then go into the perovskite line in Perrysburg. So that's why the remaining [indiscernible] capacity in Southeast Asia comes down, but it's not mixing [indiscernible] and perovskite and simply saying some of those back-end tools will now the longer be used there. They're going to go over to Perrysburg and going to be used in that 1 gigawatt pipeline for perovskite.

Our final question comes from the line of Ben Kallo from Baird.

I just have a follow-up question and then another one. Just to Joe's question, is what is the capacity after all of that's done? Because I think, Mark, you said that you lose some capacity in Vietnam and Malaysia. And so I just want to make sure that we have the volume number correct as we enter next year. And then my follow-up question is just on Topcon and your patent and what Tesla is doing and how you think about them starting manufacturing here and if that's going to violate your patent.

Yes, Ben, if you go back to the deck that we presented in February, we gave capacity and production for '26 and '27 and the assumptions at that point have not changed now. So if you look at on a production basis, we said for 2027 will be around about 19 to 20.5 gigawatts. So 19.7 I think, was the midpoint of production, and that's not changed. So if you go back to that deck, you'll see the breakout there in terms of geographical location of product.

Yes. And Ben, on the Tesla as it relates to our TOPCon patent, what we do know is that any of the TOPCon product, as you can see by the filings that we've done and the number of manufacturers who have produced TOPCon and have sold it into the U.S., effectively, all of those parties have been infringing on our IP. Okay? If Tesla chooses to go with TOPCon, my assessment would be given what I see in the market unless Tesla tries to redesign the product such that they would not infringe our IP. I guess my assumption would be there will be some form of infringement okay?

And look, I also want to make sure that it's clear that we are more than willing to work with any counterparty to engage in a commercial conversation around the licensing of our IP. We are not prohibiting that conversation, right? The issue is we just want to be paid fair value. That's also why we license the IP to talent. The talent has demonstrated a willingness to pay fair value for the technology that's enabling the product they're going to manufacture. That's fine. And we'll do that with other counterparties. And if Tesla chooses to use an IP TOPCon product, it uses our IP, then we'll center to a commercial conversation with them and happily engaged with the conversation of licensing that IP.

I mean there's nothing wrong with that, and we'd be more than happy to do that. I do think it's 1 of the challenges that Tesla is going to have to figure out what technology they go with and how do they get freed up to operate. And IP is going to be a significant challenge, at least in my mind, as it relates to TOPCon because it's a challenge that everyone else here in the U.S., I think, has realized as well. particularly as it relates to the strength of our IP, but more happy to enter into a commercial conversation with them, but they choose to do that and want to do that and we'll figure out a license range where that can work for both parties.

Look, I think Tesla getting -- establishing capability here in the U.S. market. What we've always said is we need a robust and resilient domestic supply chain completely vertically integrated beyond just [indiscernible] that's also why we're evolving towards perovskite next-generation thin films. Tesla bringing in that capacity and that capability and creating a domestic supply chain. I think it only further enhances and supports kind of the overall strategic intent around long-term energy independence and national security and having domestic supply chains, I think, are extremely valuable in both of those regards.

At this time, there are no further questions. This concludes today's call. Thank you all for attending. You may now disconnect.

Good afternoon and welcome to First Solar's Q4 and Full Year 2025 Earnings and 2026 Guidance Call. This call is being webcast live on the Investors section of First Solar's website at investor.firstsolar.com. [Operator Instructions] Please note that today's call is being recorded.

I would now like to turn the conference over to your host, Byron Jeffers, Head of Investor Relations.

Good afternoon, and thank you for joining us on today's earnings call. Joining me are our Chief Executive Officer, Mark Widmar, and our Chief Financial Officer, Alex Bradley.

During this call, we will review our 2025 results and discuss our outlook for 2026. After our prepared remarks, we'll open the line for questions. Before we begin, please note that today's discussion contains forward-looking statements, and actual results may differ materially due to risks and uncertainties. We undertake no obligation to update these statements due to new information or future events. For a discussion of factors that could cause these results to differ materially, please refer to today's press release, our SEC filings and the earnings materials available at investor.firstsolar.com.

On this call, we will also reference certain non-GAAP financial information. This non-GAAP financial information is not intended to be considered in isolation or as a substitute for financial information presented in accordance with U.S. GAAP. A reconciliation of our non-GAAP items to their respective nearest U.S. GAAP measure can be found in our earnings press release and our earnings presentation.

With that, I'll turn it over to Mark.

Good afternoon, and thank you for joining us today. Beginning on Slide 4, I will share key highlights and accomplishments from 2025. We entered the year under the new U.S. administration with a back-weighted driven profile that required sustained production to fulfill contracted commitments concentrated in the second half, amid a persistently uncertain policy and trade environment. Over the course of the year, we navigated a budgetary reconciliation process, which created the One Big Beautiful Bill Act, evolving tariff scenarios, customer negotiations and regulatory developments, including Section 232 actions, FEOC restrictions and AD/CVD investigations that are still unresolved and could ultimately prove to be either headwinds or tailwinds.

Throughout, we remain anchored to a core guiding principle and a key differentiator valued by our customers, contract certainty, both in pricing and in timely delivery. To honor obligations, we maintained sufficient capacity to fulfill international module commitments and actively pursue various contractual protections to address shifting tariff dynamics and in some cases, contract terminations. Given that backdrop, we took a disciplined selected approach to customer contracting throughout the year. That approach is proving effective. Since our last earnings call, we secured gross bookings of 2.3 gigawatts, excluding domestic India volume and 0.1 gigawatt of low-bin inventory clearance, we booked 1 gigawatt in our key U.S. utility scale market at an ASP of $0.364 per watt, inclusive of applicable adjusters.

By remaining patient, selective and opportunistic, we capitalized on demand that recognizes the differentiated value of our product and contracting structure, strengthening the forward earnings profile of our backlog and positioning us to navigate and potentially benefit from ongoing policy and trade uncertainty. We are pleased to have delivered record sales of 17.5 gigawatts of modules in 2025. Net sales of $5.2 billion were at the top end of our most recent guidance range and represented a 24% year-over-year increase.

Full year diluted EPS was within our most recent guidance range at $14.21 per share. We ended the year with $2.9 billion of gross cash and $2.4 billion of net cash, coming in above our guidance range. Our growth continued in 2025 as we advanced our U.S. capacity expansion, highlighted by initiating commercial production in Louisiana, our fifth U.S. factory.

In addition, we announced plans to onshore the finishing of Series 6 modules initiated at our international factories by adding U.S. finishing capacity with a new facility in South Carolina. We expect production from this facility to begin in Q4 of 2026 and ramp through the first half of 2027. We also advanced our Cadtel-based CuRe semiconductor platform. Following a limited commercial production run from Q4 2024 to Q1 2025, we delivered initial CuRe modules to customers in the first half of 2025.

Based on laboratory and field testing results, CuRe has demonstrated the expected advantaged energy profile driven by industry-leading temperature coefficient and long-term degradation rate with improved bifaciality. These results continue to support a disciplined factory-by-factory CuRe conversion rollout expected to begin next month, starting at our Ohio Series 6 factory.

In parallel with our Cadtel-based CuRe platform, we advanced our next-generation perovskite thin film program. Our focus remains on efficiency, energy attributes, reliability and a scalable path to high-volume, low-cost manufacturing of this potentially transformational thin film. We launched the perovskite development line at our Perrysburg campus and reached full in-line processing capabilities in Q3, marking an important step in the lab-to-fab transferability and enabling production of smaller form factor modules using anticipated manufacturing tools and integrated processes.

In late 2025, we initiated sourcing for perovskite Series 6 module form factor pilot line, which we expect to reach operational readiness in early 2027. While we made promising progress in 2025, additional work remains before more broadly scaling our perovskite program. Lastly, we continue to actively enforce our intellectual property rights, including our TOPCon patents in the U.S. Notably, in Q4, the U.S. Patent and Trademark Office denied 3 separate petitions filed by foreign headquartered manufacturers that sought to invalidate aspects of our TOPCon portfolio. This outcome reinforces our confidence in the strength of our patent portfolio.

I will now turn the call over to Alex to discuss our most recent shipments and booking activities as well as our Q4 and full year 2025 results.

Thanks, Mark. Beginning on Slide 5. As of December 31, 2024, our contracted backlog totaled 68.5 gigawatts valued at $20.5 billion or approximately $0.299 per watt. For the full year 2025, we sold 17.5 gigawatts of modules, secured 7.4 gigawatts of gross bookings and recorded 8.3 gigawatts of debookings, primarily due to our termination of contracts as a result of contract breaches by customers, resulting in net full year debookings of 0.9 gigawatts.

We ended the year with a contracted backlog of 50.1 gigawatts valued at $15 billion. As a reminder, the contracted -- backlog reflects the base ASP. A significant portion of our existing contracted backlog includes pricing adjusters that may increase the base ASP contingent on achieving specific milestones within our technology road map and manufacturing replication plan. At year-end, approximately 23.2 gigawatts of contracted volume included these adjusters, which we estimate could generate up to an additional $0.6 billion or approximately $0.03 per watt, the majority of which will be recognized in 2027, 2028.

Turning to the P&L on Slide 6. Q4 net sales were $1.7 billion, a $0.1 billion increase sequentially. Full year net sales were $5.2 billion, a $1 billion increase year-over-year, driven primarily by a 24% increase in module volume. Gross margin in Q4 was 40%, an increase from 38% in the prior quarter. The increase was driven by a higher mix of U.S. manufactured modules benefiting from Section 45X tax credits, lower nonstandard freight charges due to reduced international shipments and the resolution of the glass supply chain disruption experienced in Q3 at our Alabama facility.

These benefits were partially offset by ramp and underutilization costs for Louisiana, a higher proportion of sales into the India market and the termination amounts recognized in the third quarter related to the breach of contracts by affiliates of BP. Full year 2025 gross margin was 41%, a decrease from 44% in the prior year. The decline was primarily driven by tariff costs as well as the impact of tariffs exacerbating warehousing expense associated with a back-weighted revenue profile, detention and demurrage, partially driven by supply-demand imbalances following certain contract terminations due to customer default and underutilization from the curtailment of our Series 6 international facilities.

These headwinds were partially offset by $1.6 billion of Section 45X tax credits recognized in 2025 relative to $1 billion in 2024, driven by a higher mix of U.S. manufactured module volumes sold. As an update on warranty-related matters, we've resolved certain claims and have continued to advance negotiations with additional customers regarding warranty claims for select Series 7 modules produced prior to 2025. Based on our settlement experience, the estimated number of effective modules and projected remediation costs, we believe a reasonable estimate of potential future losses will range from approximately $35 million to $75 million.

Within this range, we recorded a specific warranty liability of $50 million, representing our best estimate of the expected impact associated with this issue. We're aware of certain statements, including in recent counterclaim filings by affiliates of BP relating to overall PV plant underperformance. While we will not comment on existing litigation, we do encourage a review of our initial complaint filed last year as well as our answer to those claims in our motion to dismiss filed earlier this month. As it relates to overall PV project output, we would note that solar plant performance relative to expectation is influenced by a broad set of environmental, design, operational and grid-related factors.

These include, but are not limited to, third-party prediction modeling, weather variability, terrain variability, local microclimates, shading and soiling, procurement and design decisions relating to trackers, inverters, transformers and other plant components, design and construction parameters, including EPC quality, DC and AC system design, construction and handling and operational factors, including tracker algorithm design and fidelity, open circuit conditions and overall O&M scope and quality.

In short, the solar plant's performance reflects its total environment system design. As we've consistently stated, First Solar fully stands behind its module warranty obligations. To the extent a customer has a valid module warranty claim, we remain ready, willing and able to perform our responsibilities pursuant to the procedures agreed to in our warranty. SG&A, R&D and production start-up expense totaled $117 million in Q4, a decrease of approximately $27 million relative to the prior quarter. The decrease was primarily driven by the reduction of start-up costs associated with the Louisiana facility commencing commercial operations.

For the full year 2025, operating expenses were $523 million, an increase of $59 million year-over-year. This includes a $42 million increase in R&D expense, driven by higher depreciation, maintenance and utility costs associated with our research facilities as well as increased headcount and compensation. SG&A increased by $15 million, driven primarily by higher allowance for credit losses on aged receivable balances and supplier loans. Our fourth quarter operating income was $548 million, which included depreciation, amortization and accretion of $141 million, ramp and underutilization costs of $29 million, excluding depreciation and production start-up expense of $1 million and share-based compensation expense of $3 million.

For the full year 2025, our operating income was $1.6 billion, which included depreciation, amortization and accretion of $529 million, ramp and underutilization costs of $140 million, production start-up expense of $86 million and share-based compensation expense of $19 million. Interest income, interest expense, other income and foreign currency losses totaled $3 million in income in Q4 and $16 million of expense for the full year.

Income tax expense for the fourth quarter was $30 million compared to a tax expense of $4 million in the third quarter. And this quarter-over-quarter increase in tax expense was primarily a function of Q3 benefits, including a $20 million discrete tax benefit associated with the acceptance of a filing position on amended tax returns in a foreign jurisdiction and incremental share-based compensation benefits recorded in the prior quarter. Recorded full year income tax expense of $53 million. Q4 earnings per diluted share were $4.84 compared to $4.24 in the previous quarter. For the full year 2025, earnings per diluted share were $14.21 compared to $12.02 in 2024 and within our guidance range.

Turning to Slide 7, I'll cover select balance sheet items and summary cash flow information. The aggregate balance of our cash, cash equivalents, restricted cash, restricted cash equivalents and marketable securities was $2.9 billion at year-end, an increase of $0.8 billion sequentially and $1.1 billion year-over-year. Both the sequential and full year increases in gross cash were driven primarily by proceeds from the sale of Section 45X tax credits generated during the year and positive operating cash flows, partially offset by capital expenditures for our Louisiana facility.

We monetized $0.8 billion of 2025 Section 45X tax credits in the fourth quarter and $1.4 billion during the full year. Notably, in January 2026, we also received $118 million for 2024 Section 45X tax credits where we elected a direct pay option in our 2024 tax return filed in October of 2025. The sale transactions highlight the liquidity of the Section 45X tax credit sale market and the IRS refund provides insight into direct pay election turnaround times, providing additional visibility and flexibility to optimize credit monetization and manage overall liquidity.

Accounts receivable and inventory decreased both sequentially and relative to the prior year, reflecting improved customer collections and higher volumes sold. Capital expenditures were $172 million in the fourth quarter compared to $204 million in the third quarter. Full year 2025 CapEx was $870 million compared to $1.5 billion in 2024. Our year-end net cash position was $2.4 billion, an increase of $0.9 billion from the prior quarter and an increase of $1.2 billion from the prior year.

I'll now turn the call back to Mark who will provide an update on market conditions, policy and technology.

All right. Thank you. Turning to Slide 8. In 2025, the policy and trade environment remain complex. While we are experiencing significant direct and indirect tariff impacts, in our view, on balance, the environment is net favorable for First Solar, an example of genuine long-standing U.S.-based solar manufacturing. In contrast, in our view, headwinds beyond reciprocal tariffs and commodity cost increases continue to build for the crystalline silicon industry, a combination of tighter trade enforcement, potential retroactive tariffs, pending Section 232 actions, expanding foreign entities of concerns or FEOC restrictions and greater intellectual property enforcement is increasing cost, timing and compliance risk for developers relying on crystalline silicon products with ties to China.

With respect to trade, it is notable that the Trump administration has withdrawn its appeal against the U.S. Court of international trade ruling and the auction litigation requiring the retroactive collection of previous suspended AD/CVD tariffs. If this ruling is maintained, which appears increasingly likely, mounting contingent liabilities for AD/CVD duties associated with this unlawful 2-year moratorium could represent an as of yet unrealized material financial impact on those foreign producers that relied on it.

In fact, one recent industry publication noted that "U.S. customs is suggesting that no panels that came in during the moratorium qualified for the moratorium." In support of true domestic manufacturing, we are encouraged by interim treasury guidance issued earlier this month that, in our view, clearly signaled the administration's intent to address gamesmanship by Chinese tied solar manufacturers seeking to evade U.S. regulations and benefit from tax incentives by artificially shuffling ownership stakes, voting rights or IP licensing with the intent of evading FEOC status.

We anticipate that the forthcoming FEOC restrictions will be aligned with the legislative intent of prohibiting access to tax credits for entities with sudden corporate restructuring lacking business purpose. We also commend Commerce's preliminary CVD determinations issued earlier today as part of the broader Solar 4 AD/CVD investigation into Laos, India and Indonesia, which reflect subsidy rates of approximately 81%, 126% and 104%, respectively.

Note, these preliminary CVD rates do not include preliminary antidumping rates, which will be additive and are expected to be determined in April. It is expected that final aggregate AD/CVD rates will be determined by the end of September. We expect that the final AD/CVD duties applied in Solar 4 will again support a level playing field against the illegal and unfair trade practices of Chinese headquartered and other crystalline silicon manufacturers who strategically evade U.S. trade laws.

Finally, IP enforcement must be considered within the overall legal framework that includes these recent legislative and regulatory headwinds confronting the crystalline silicon industry. Earlier today, we filed a petition with the U.S. International Trade Commission, or ITC, against 10 groups of foreign headquartered manufacturers that we believe are producing products infringing on one of our U.S. TOPCon patents. Note, this is a separate action from our 3 actions seeking monetary damages against affiliates of Adani, Canadian Solar and Jinko in the U.S. District Court.

If the ITC institutes an investigation based on our complaint, we expect that the matter would be decided in approximately 18 months. If our case is successful, the ITC may issue a general exclusion order preventing the importation of infringing TOPCon products made by foreign entities or an alternative may issue a limited exclusion order preventing the importation of infringing TOPCon products by the entities named in our complaint.

In addition, the ITC may issue a cease and desist order preventing the sale of infringing TOPCon products currently in the United States. Note, the IP-related headwinds confronting the crystalline silicon industry is reflected not just by First Solar's recent decrease and continued enforcement efforts, but also by the recently reported $236 million settlement entered in between Maxeon and Aiko, just days before a JUVE Patent court was due to decide their patent dispute.

In summary, the policy and trade environment, together with sustained intellectual property enforcement across the industry have continued to generate mounting uncertainties for U.S. developers that are dependent on suppliers tethered to China tied supply and/or IP. On the top of the technology, turning to Slide 9. Our strategy remains anchored in a simple premise. Customers ultimately buy lifetime energy, not just nameplate efficiency. And our road map is designed to optimize the balance of efficiency, energy yield and cost while leveraging our industry-leading thin film expertise.

We continue to believe the next step change in solar will be enabled by thin film platforms such as perovskites. We are well aware that many in the industry are seeking to crack the code of this potential next generation of advanced thin film technology. However, we believe the winner of the perovskite race will also need to have the ability to manufacture the product cost competitively in a high-volume manufacturing environment. As the world's leader in thin-film PV technology and the world's only manufacturer of thin films at scale, we believe we are uniquely positioned to advance this prospective device given our nearly 3 decades of not only thin film R&D learnings, but high-volume thin film manufacturing experience.

Our technology strategy continues to be primarily concentrated on 2 core thin film focus pillars. Firstly, we are executing a disciplined phase-gate introduction of CuRe, responsibly bringing this new technology to market with proven laboratory results and expanding field validation. Consistent with our prior outlook, we expect to permanently convert the Ohio lead line to CuRe in Q1, providing a pathway to enhance the energy attributes and competitiveness of our Series 6 platform and then rolling out these enhancements to our Series 7 platform at successive factories.

Executing CuRe remains strategically important because it is designed to enhance the attributes that translate into lifetime energy, including improved temperature response and degradation behavior, which are highly valued by utility scale customers. When adding the improved energy attributes of CuRe to the current energy advantages of thin-film CdTe, such as superior spectral and shading response, CuRe can deliver up to 8% more lifetime specific energy yield than crystalline silicon TopCon technology in the markets we serve.

Our second pillar, perovskites, is a key part of our effort to develop next-generation thin film semiconductors that can be deployed at commercial scale in both our traditional utility scale markets while potentially expanding our addressable market segments. To date, we have achieved reliability results we believe are comparable with best-in-class R&D efforts while continuing to advance efficiency and stability, 2 of the industry's key hurdles to scaling perovskite technology. A major enabler of these efforts is our dedicated perovskite development line in Ohio.

As announced prior to the beginning of the call, we have entered into an agreement with Oxford PV, the holder of what we believe is the most fundamental portfolio of perovskite-related patents. Under the terms of this agreement, Oxford PV will license to us on a nonexclusive basis, its existing issued and currently pending patent applications. We believe that this agreement will advance our freedom to develop, manufacture and sell crystalline silicon-free perovskite-based semiconductor modules in the U.S. utility, commercial and residential markets.

I will now turn the call back over to Alex, who will discuss our 2026 outlook and guidance.

Thanks, Mark. Before turning to our financial guidance, I want to briefly reiterate our approach to managing the business amid a dynamic market policy and trade environment. We entered 2026 with a backlog of 50.1 gigawatts despite taking a highly selective approach to bookings over the past 2 years. We expect to continue this strategy in 2026 as we await the outcome of and impact on forward module demand and pricing from the numerous political, regulatory and legal matters discussed earlier in the call.

We believe our market position with non-FEOC and high U.S. content supply on our proprietary thin film platform remains long-term advantage and our existing contracted backlog relative to our production capacity provides us with the ability and flexibility to be patient. Regarding 2026 U.S. deliveries, many of our customers continue to face both regulatory and commercial challenges, including federal permitting approval delays. As we previously stated, we will continue to work with customers to accommodate schedule shifts where possible, consistent with our philosophy of supporting our long-term partnerships even while we are not contractually required to do so.

We entered 2026 with a fully allocated position for our U.S. production. Given tariff uncertainty, our India production is assumed to be sold into the India domestic market. We'll continue to monitor opportunities to export products into the U.S. where it is margin accretive to do so. Our guidance assumes our India facility operating at full capacity with the ability to flex production as needed through the year in response to changes in demand signals. Demand for our Series 6 international products produced in Malaysia and Vietnam remains constrained. Our decision in Q4 of 2025 to establish a new finishing line in the U.S. allows us to make use of a portion of the front end of these Southeast Asian facilities, optimizing freight, tariffs and domestic content for the sale of incremental products into the U.S. domestic market.

We intend to run our remaining end-to-end capacity in Malaysia and Vietnam at low utilization rates this year despite the financial impact of doing so, maintaining a near-term option to increase throughput should catalysts such as the political and regulatory matters discussed earlier, drive incremental profitable demand.

Slide 10 shows our capacity and forecast production for 2026 and into 2027. Nameplate capacity reflects current output entitlement at full scale throughput and yield with downtime solely for planned maintenance. Production reflects nameplate capacity adjusted for factory ramp, other downtime, including the technology and tool upgrades and any planned reduction in throughput, including due to market demand. As it relates to technology, as previously noted, we expect to recommence running Series 6 CuRe products in Perrysburg this quarter. Assumed in our 2026 production forecast for India is downtime associated with tool upgrades with the intent of beginning CuRe production on our first Series 7 line in India in early 2027, followed by the remainder of the Series 7 fleet thereafter.

Additionally, our Southeast Asian capacity has reduced significantly in both 2026 and 2027 due to the removal of tools destined for our U.S. finishing line as well as the reuse in our [indiscernible] site development work. By 2027, U.S. finishing capacity is forecast to be 3.5 gigawatts with remaining Southeast Asia capacity of 1.8 gigawatts for fully finished international Series 6 modules.

With projected U.S. nameplate capacity of 14.9 gigawatts in 2026, growing to 17.1 gigawatts in 2027 with the scaling of Louisiana and South Carolina. We expect global nameplate capacity of 19 gigawatts in 2026 and 22.1 gigawatts in 2027. And note, we expect U.S. nameplate to continue to increase as we drive technology throughput and yield improvements. In terms of production, as previously noted, we expect significant underutilization of our international Series 6 facilities. India is assumed to run high throughput with the ability to flex production as needed based on local demand and international sale options. With forecasted U.S. production of 13 to 13.3 gigawatts this year and 14.9 to 16.1 gigawatts next year, this results in total forecasted production of 16.5 to 17.5 gigawatts in 2026 and 18.9 to 20.5 gigawatts in 2027.

Turning to Slide 11 and other assumptions embedded within our guidance today. Expected volumes sold of 17 to 18.2 gigawatts is above forecast production as we reduce inventory levels by year-end. We forecasted ASP of approximately $0.308 per watt for volumes sold in the U.S., slightly above our contracted backlog. This includes certain freight, tariff and commodity recovery and technology upside. Now we expect limited ASP upside from CuRe sales in 2026, largely as a function of contractual notification deadlines relative to the timing of the decision to recommence CuRe production.

Combined with India domestic sales, the majority of which we expect to book and deliver within the next year, we forecast a global ASP recognized of approximately $0.287 per watt. Cost per watt sold is forecast to remain relatively flat year-over-year at approximately $0.267 per watt, which includes the impact of tariffs and the impact of reshoring U.S. manufacturing, largely recognized through ramp and underutilization expense and excludes the benefit of Section 45X credits generated by domestic manufacturing.

Including the benefit of Section 45X credits, cost per watt sold is projected to be down approximately $0.03 per watt year-over-year. Cost watt released from inventory is expected to increase approximately $0.02 per watt year-over-year. Approximately half of this increase is forecast to come from a mix shift as both Southeast Asia production reduces and Louisiana and Alabama increases and the other half to come as a result of multiple factors, including increases in tariff costs, increases in core bill of material costs, increases in utility rates and downtime for technology upgrades.

Period costs are expected to decrease by an equivalent approximately $0.02 per watt from a combination of lower standard sales rate due to greater domestic mix shift, the effective elimination of nonstandard freight charges as a function of reduced international product imports, reduced warehousing costs and other period cost reductions.

To provide some more color, we forecast total net tariff cost impact across bill of material and finished goods imports recognized in both cost per watt produced and period costs towards cost per watt sold of $155 million to $175 million. Net of expected contractual recoveries on finished goods sold, we expect a total tariff impact of $125 million to $135 million. This assumes a Section 122 tariff in place for 150 days at 15%, impacting all bill of materials, works in progress and finished goods imports in that time period.

In addition, certain commodities, including aluminum, are subject to long-lasting higher rate Section 232 tariffs. And note that we recognize the P&L impact of tariffs at the time of product sale, our total tariff impact in 2026 reflects higher tariff rates paid on bill of materials, works in progress and finished goods imports incurred prior to the recent Supreme Court decision. Tariffs also indirectly due to underlying commodity cost pressure for our variable U.S. bill of material, including relating to aluminum, steel, glass, interlayer targets and spares.

In addition, electricity rate hikes increased fixed costs. These cost increases are not contractually recoverable from our customers. Sales rate is forecast to be approximately $0.014 per watt in 2026. Warehousing-related costs of approximately $200 million is down from 2025, but remains high in part of a function of underutilization of space driven by our forecasted Southeast Asian curtailment. Beginning in 2027, we expect to reduce warehouse costs to a longer-term run rate of approximately $100 million per year. Forecast ramp and underutilization expenses of $115 million to $155 million are a function of curtailing Malaysia and Vietnam capacity as well as the ramp of our U.S. finishing line.

Setup cost is expected to be $110 million to $120 million, driven primarily by the ongoing depreciation and logistics associated with idled finishing equipment in transit as well as certain tariff-related costs incurred to import that equipment and warehousing such equipment for use at our new South Carolina facility. And finally, note on capital structure. Our strong balance sheet remains a strategic differentiator. It allows us to navigate periods of volatility, including near-term supply and demand imbalances in certain international markets, while continuing to support R&D investment, technology capital spend and capacity growth, including localizing the U.S. solar supply chain in support of energy security and national policy objectives.

We ended 2025 in a strong liquidity position and have since enhanced our financial flexibility by entering into a new $1.5 billion senior unsecured revolving credit facility, improved commercial terms, greater flexibility and more relaxed covenants. In 2026, we intend to fund CapEx through cash on hand and operating cash flow. We plan to prepay the remaining balances outstanding under our India credit facilities, including our loan with the DFC ahead of its scheduled maturity.

Doing so enables us to optimize our jurisdictional capital structure, increase the capacity of our local working capital facilities while reducing exposure to India rupee volatility-related hedging costs. Separately, while not assumed in our guidance, potential monetization of 2026 Section 45X tax credits provides additional liquidity optionality. Our capital allocation priorities remain unchanged. Firstly, we prioritize maintaining a resilient working capital reserve of approximately $1.5 billion to $2 billion to account for industry cyclicality and uncertainty and short-term supply and demand imbalances.

Secondly, we deploy cash to fund growth and replicate technology improvements across the fleet. Thirdly, we invest in innovation through R&D and targeted capital investments as well as strategic enablers such as licensing arrangements to advance our technology road map, including perovskite optionality. Fourthly, we'll consider M&A, which we're actively evaluating to pursue complementary technology adjacent opportunities to reinforce our strategic differentiation. As we near the conclusion of recent years of sustained high CapEx associated with manufacturing capacity growth, we will evaluate applying cash generation in excess of the above capital priorities to share repurchases. We'll provide an update on this later in the year, pending clarity around certain factors, including policy catalysts, realization of new bookings volume and any decisions around 2026 Section 45X tax credit monetization.

I'll now cover the full year 2026 guidance ranges on Slide 12. Our net sales guidance of between $4.9 billion and $5.2 billion. Gross margin is expected to be between $2.5 billion and $2.6 billion or approximately 49.5%, which includes $2.1 billion to $2.19 billion of Section 45X tax credits and $115 million to $155 million of ramp and underutilization costs. SG&A expense is expected to be between $215 million and $225 million and R&D expense between $285 million and $290 million. The primary driver for our increase is due to higher investment in advanced research, including expanded perovskite innovation center activity and planned headcount additions.

SG&A and R&D expense combined is expected to total $500 million to $515 million and total operating expenses, which includes $110 million to $120 million of production start-up expense, are expected to be between $610 million and $635 million. Included within our R&D expense guide for 2026 are approximately $100 million of costs associated with perovskite development. Going forward, we intend to guide on an adjusted EBITDA basis, which we believe provides the clearest view of our underlying operating performance, enhances comparability across periods.

We forecast full year adjusted EBITDA of $2.6 billion to $2.8 billion. From a first quarter earnings cadence perspective, we expect module sales of 3.4 to 4 gigawatts, Section 45X tax credits of $330 million to $400 million, resulting in adjusted EBITDA of between $400 million and $500 million. Capital expenditures in 2026 are forecast to range from $0.8 billion to $1 billion. Approximately half of the spend will support capacity expansion, primarily the South Carolina finishing line and the Louisiana plant. Remaining spend is expected to be split evenly between CuRe in India and R&D technology replication and maintenance.

We expect to end 2026 with gross and net cash balances between $1.7 billion and $2.3 billion and assume a full repayment of our India credit facility with the U.S. International Development Finance Corporation by June 30, 2026.

With that, we conclude our prepared remarks and open the call for questions. Operator?

[Operator Instructions] Our first question comes from the line of Brian Lee with Goldman Sachs & Co.

I guess just on the ASP front, Mark, you mentioned the $0.364 per watt, including adders for the U.S. bookings this quarter. How much did the adders add there? And then is that sort of the level of entitlement, $0.36 and above that you would expect for U.S. bookings through the rest of the year? Maybe can you comment on visibility you have on the pricing environment from this point forward? And then just secondarily, on the gross margins, I'd be curious, I mean, this is implying kind of a 10% component gross margin in the guidance, even if I factor out underutilization and the 45x credit. So I know there's a lot of moving pieces, Alex, but when do you guys kind of get back to high teens, 20% type of gross margin for components the way you were at in 2024? And kind of what are some of the big bridges to get back there?

All right. I'll do the ASP conversation and then Alex, obviously take the gross margin. But -- so there's about, call it, $0.025 to $0.03 of the value of the adder in terms of the ASP, that $0.34 -- or excuse me, $0.364. So that's about the number. So the CuRe attributes will give you close to 3%. Not everything that was booked in that 36.4% actually had the adder. So if you sort of broke it out a little bit, you potentially going to see slightly higher prices for the CuRe attributes for the product that we booked with the CuRe attributes.

But look, I think the entitlement wise, I mean, I feel good about the pricing that we're at now, but I think there's some more catalysts that could still happen. I think some are leaning in with all the uncertainty that we referenced on the call in terms of what happens with 232 and what window are they trying to book into given the constraints with FEOC. And I think if we see -- continue to see more momentum there, we just obviously saw the Solar 4 announcements, which obviously are going to address imports coming in from those 3 countries that we referenced. So there's potentially more tailwinds depending on how they play out that could support even better pricing as we move forward. But I think visibility-wise, I think that's kind of a good indication of where we expect the market pricing to be relative to the CuRe technology.

Yes. Brian, on the gross margin, so if you back out the 45X, it's about a 7% gross margin. And if you think about that against the $5 billion revenue profile, it's about $350 million of gross margin. So if you try and walk that back up to, call it, the 20% number we've talked about before. This year, we've got about $165 million of tariffs sitting in there and about $135 million on utilization. So you have about $300 million or so there, about 6 points of margin. You've got about $200 million of warehousing. I think we said in the prepared remarks, that will come down to about $100 million on a run rate basis next year. So pick up another $100 million in there, another 2 points of margin. And then if you look at the adjusted, we talked about $600 million of the backlog mostly recognized in 2027, 2028. So you take $300 million across each of those years, that's another 6 points of margin. That kind of walks you back up to around about $1 billion of gross margin and a 20% number.

So tariff is obviously still in that. That's an impact that we're still wrestling with. But I'd also say you've got incremental volume coming in next year, which is not factored into that and the contribution margin benefit of that even ex IRA is going to be meaningful. I'd also say it's -- we're trying to look at this on an ex IRA basis, but it's challenging to do that to some degree because we are adding incremental costs, both by U.S. manufacturing and also by the mix shift to bringing more production into the U.S.

And we're doing that, and that allows us and enables us to capture that 45X credit. So I know we're looking on an ex IRA basis. But if you think about it with the IRA included with that 45X yes, the core also is 7% this year and the walkup that we talked about just now, but you've got 43 points of IRA. So the gross margin on a GAAP basis will be 50% this year relative to 41% last year, which is the highest we've seen. So obviously, we want to keep growing the core underlying non-IRA, and there's a path there, as I said, to take you back up to that $20 million and we're about $1 billion at that $5 billion revenue profile. But you can't ignore the 2 and the interconnectedness of the fact that we are reshoring capacity, and that does come at a cost, both direct cost and mix shift cost that enables us to capture those 45X benefits.

Yes. And Brian, maybe I want to add a couple of things to. One back to on the pricing environment, I also want to make sure this is clear is that, that product or that ASP is also reflective of largely influenced by a domestic content product, right? Now what we do a lot of times is it's agreed to around some number of points, which allow us to bring in some amount of international volume, assuming the tariff rates are amenable such that we can blend international. But if we ended up doing a straight up international deal, for example, with some of the capacity we have available with our Malaysia, Vietnam facilities, I would expect that price point to be closer to $0.30.

So I just want to make sure that, that's clear. So if we actually end up booking some of that international volume as we go forward, that you may see a lower ASP there because of that dynamic. I want to make sure that's clear. And then the other one, just to Alex's point around tariff. The other thing that's kind of creating some headwind for us this year is -- and we've mentioned this before, I mean, there is insufficient glass supply in the U.S. And we've been working to bring in basically brownfield and mothballed facilities and getting them up and operational. But to support the scaling that we have right now, we are bringing in some glass internationally and using it in our U.S. production, and that is creating a cost headwind. My inbound freight costs are higher. I'm bearing the cost of the tariffs. And what we would expect to do as we scale up our supply chain here in the U.S., which we're in the midst of doing, we'll see less of a dependency on that import of that glass, which will also help us a little bit on our gross margin profile.

Your next question comes from the line of Julien Dumoulin-Smith with Jefferies LLC.

Look, if I can ask you, first off, just to reconcile on the volumes produced versus sold, -- can you comment a little bit about just what you're seeing here of late? And then separately, can you comment a little bit about what you're actually seeing in terms of sell-through on volumes out of Asia? I know your prepared comments had some nuance on this, but just elaborate a little bit about what both Southeast Asia/India are assuming here in '26 and what you're seeing preliminarily '27 as well.

Yes. So if you go to the slides, we give you a walk on what we're expecting to produce versus sell. The delta between the 2 is about 700 megawatts coming out of inventory. So that's the gap you're seeing between produced volume and sold volume. And we gave you the U.S. number for sold. You can see it in there. It's $0.29.33 [indiscernible] It's in the script, I'll come back to you. But yes, the delta is coming out of inventory. And what was your second question, Julien?

I think part of your question was on sell-through of Southeast Asia, I think, is what you...

Yes, on sell-through. What are they [indiscernible]

So yes, let's talk through that. So India is going to produce, call it, 3 gigawatts or so this year and will be sold into the India market. We're actually going to have a really strong quarter. We had a really strong Q4, and we'll have a strong Q1 in India. So demand in India is strong. So we got 3 gigawatts or so, a little bit north of that because we got some inventory sell-through that we'll sell in 2026 in India domestically manufactured, sold into the India market, okay?

Now on the Southeast Asia, those factories are kind of running 20% or so. I mean they're extremely underutilized. And what Alex said in his prepared remarks is that we're looking at this almost as option value. Now some of that capacity will be fully utilized once we get our South Carolina facility up because the front-end capacity for, call it, half of the Southeast Asia manufacturing will come to the U.S. So that will solve a piece of that underutilization.

The other half is going to continue to be underutilized at a very low utilization rate. But what we're looking at this is really an option to allow some of these potential tailwinds around 232 and other things to play itself out to see what the impact of that could be to create more demand for those international operations. We also, I think, as we indicated in our last call, we are trying to work with a couple of counterparties on potentially meaningful volume offtake for those international production for that international production. But that's all still in the works. It's still going to be largely tethered back to what happens with the policy environment.

But we are incurring significant underutilization and cost headwinds because of what we're trying to do right now is figure out, let's create an option, let's evaluate what we continue to see in the market. And we've been sort of wearing this for over a year now. One -- it was about a year ago when these tariffs came in place, and we started to throttle down kind of towards the end of Q2, beginning of Q3, and we've kind of run at a very low utilization rate to try to sort of buy some time to see how these tariffs ultimately get played out.

Your next question comes from the line of Mark Strouse with JPMorgan...

Within the last couple of months, there was an individual with a vast amount of resources that is talking about ramping up supply of U.S.-based solar panel production over the coming years. Just curious if that is having any real impact on the conversations you're having with your customers, especially if you look out to later this decade? And I've got a quick follow-up.

Yes. Look, obviously, very much aware of the announcement and the ambitions. From my understanding, a lot of that is not necessarily focused for kind of our utility scale market that we're primarily focused on. I think it's primarily to be captive for their own consumption for their own programs that they're envisioning. It's also out in our horizon. And I think there's also a pretty strong realization of some pretty significant challenges to try to get to that type of scale. And if you're really thinking about this as fully vertically integrated all the way from polysilicon forward, you not only are trying to solve the constraint at, let's say, the cell level, you've got to think through how do you sell the wafer and how do you think through the polysilicon.

And the capital investment to try to move all that forward is going to be pretty overwhelming. And the technical aspects around it as well of understanding freedom to operate issues. We've already talked about there are IP infringements all over the place within the crystalline silicon world, and everybody is going to stand up and try to protect their IP where it makes sense. So it hasn't really impacted our conversations yet. I think if they got to a realization where you started to see sites being announced, actual equipment being purchased, operations being commenced, then I think you could -- maybe it starts to sort of inform our customers' views of other thoughts and ideas. But as of right now, it's having very little impact.

Mark, just one other comment around that. As you're well aware, the biggest constraint that we're seeing in the market for hyperscalers right now is access to power. And as we've just gone through the process of looking to site our new finishing line, the biggest constraint we faced around that was land with available power to connect. So again, Mark mentioned, you've got kind of capital constraint, which even that could be overcome, IP constraints, just knowledge and know-how. You've also got the constraint of how you would power these facilities if you're trying to build to that scale, you'd be in the same constraint that the hyperscalers have today.

Okay. That makes sense. And then I just want to ask a clarifying follow-up to Brian Lee's question earlier. Mark, when you said the $0.364 is for the domestic content, are you saying that, that is the blended rate? So it would be a higher amount than that for the U.S. domestic content average with whatever, $0.30 for international? Or is that $36.4 how we should be thinking about the domestic content portion?

Yes. So the way we -- and I tried to mention this a little bit last time. The way we actually price is we negotiate with the customer some number of points that they need, given their the already procured strategy around -- procurement strategy around the tracker in particular, and then what are the incremental points that they need relative to the window that they plan on putting the project into service. And so we will negotiate a points construct, which then gives us the optionality to blend whatever percentage we can internationally.

And we're in a pretty good balance between the domestic S7 and the domestic -- excuse me, international S7. We can optimize there pretty well, assuming we choose not to sell into the Indian market. Where we're a little bit more of misaligned is really with the only having, call it, 3 gigawatts of capacity for S6 in the U.S. and then try to match that up with 7 gigawatts of international production, which makes it harder because you can't really blend that much international to achieve, let's say, a 30- or 40- or 50-point requirement that a customer has.

Now moving some of that finishing capacity in the U.S. also brings in some domestic content. So that helps move that equation a little bit. But what I was trying to say that if you have no domestic content at all embedded into a blended price construct, you're going to see a much lower. So if I go out and say, okay, I want a customer to buy 500 megawatts of international only with no domestic content value. And you may see some of that in some customers that are doing PTC project, as an example, because the uplift on the PTC is not as meaningful as it is on the ITC. Those prices are going to be lower. And I just want to make sure that's clear and that's understood that you could see a delta. I said $0.30 or somewhere in that range. If you're selling just a pure international S6 product into the U.S. market, you would see a lower ASP clearing price.

Your next question comes from the line of Philip Shen with ROTH Capital Partners.

First one, just was wondering if you could give us a little more color on why no EPS guide for '26. And then secondly, in terms of the ASP implied for the 2026 guide, it seems like the U.S. ASP might be a little bit low. I think in 2025, it was closer to $0.324, but the implied U.S. ASP in the '26 guide seems to be closer to $0.308. So I was wondering if you might be able to give some color on that. And then finally, on Oxford PV, can you share what kinds of efficiencies you're able to generate? What are you seeing in your test modules, if any? And then what's your sense of timing as to when commercial volumes could actually ramp?

Yes. So on the EPS piece, we're moving to guiding to EBITDA. We think it gives a better view of operational performance. It's better comparability year-over-year. And especially a year like this year, we've got significant costs associated with both underutilization of our Southeast Asia facilities as we're deliberately curtailing those, as Mark mentioned, waiting for an option around whether there's profitable capacity or profitable production to be had there to serve the market.

And we've got significant start-up costs as we're moving that equipment from Southeast Asia to the U.S. And so historically, we haven't done that around start-up and ramp production, but this is a little different where effectively taking tools we already have in the ground and idling them for a significant period of time. So I think it makes it more comparable. On the tax side, also, we've got potential challenges around Pillar 2 this year, which will make the tax number very noisy going down to EPS. So on a kind of core non-Pillar 2 basis, we expect very low tax expense for the year. But because the agreement has not yet been finalized, there's a potential chance we'll have to accrue significant Pillar 2 expense, which ultimately we don't believe will ever get realized on a cash basis. But until those agreements are finalized, we'd have to accrue that. So we think it helps us with comparability as well.

On the ASP side, so I think your number is right, $0.308. That's -- if you look at what's in the backlog, it's around $0.30. So it's about what's in the backlog coming out of that ASP with a little bit of uplift relative to some adjusters we get around freight, around commodities. And then there's a little bit of upside on the tech side as well there. If you think about -- we mentioned on the adjustments going forward, there's about $600 million or about $0.03 a watt on average of adjusted value sitting aligned with the CuRe platform, and most of that we said will be realized '27, '28, which is when we'll start having much more CuRe product available to us. So this year, we have limited CuRe production. And given the timing of our decision to run that CuRe relative to the notification timings in the contracts, we're going to see very limited upside from that this year. So that's why you're seeing that U.S. ASP around 30.8%.

I think the other thing, Phil, when you're looking at your year-on-year, you have to remember that last year, as we realized throughout the year, a handful of terminations, obviously, the largest one being the Lightsource BP termination, which obviously impacts revenue, but you'd have to normalize, pull that out of your top line revenue and then do your math from that standpoint. But that clearly provided some incremental uplift to the reported ASP last year because of some of those terminations.

As it relates to the Oxford PV, or just our perovskite program. I mean, basically, what we're doing right now, Phil, is we have a development line where we're doing small form factor modules. So thinking they're like 60 centimeters by 20 centimeters or so. So these are actual fully functional small form factor modules that we're producing in an actual integrated production process and doing all of our kind of conversion of some of the efforts that we do in our advanced research, whether it's in California, whether it's in Sweden, and we're doing all that testing within our development line.

And if you look at our -- as we said in the prepared remarks, if you look at our efficiencies and what we're seeing in stability, we're best-in-class from many, many ways, right, many different dimensions. So I feel good about where we are from a program standpoint. But there are still fundamental challenges that have to be addressed. And we are very good at understanding the nuances of thin films, right? We understand the issue of [ metal ] stability, right? We understand the impact of encapsulating the thin film, right, to protect the thin film. So we have a number of what have been identified as best-in-class, in particular, Chinese perovskite product.

And if you put those out into a test field as we have, first off, they'll even tell you in their literature that they give you do not expose the module to open circuit, which means effectively once you install it, it has to be immediately energized. And if you choose to expose it to open circuit, it degrades almost instantaneously in some cases. There's others that because their encapsulation is so poor that the film just starts to, over time, effectively pull apart, right? So there's a huge significant delamination that happens with the film and it can sustain itself over extended periods of time.

And so there's a lot of many factors -- many, many factors that factor into how do you commercialize the product. We are working on all of those dimensions, right? So our view is we need to think about the development, not just to drive the efficiency, also drive the attributes to a point where they're competitive and then to manufacture that in a way that creates an enduring product that can be into the field for 30-plus years and perform at attributes that are relatively close to the current thin film technology that we have, and then we can do that all in HVM and do that in a way that we can do it cost competitively.

So there's many different things that we're working on in that regard, Phil. So I would just say we're pleased with where we are. We are -- our next phase is we will be investing. We've already started our procurement process to put it together a pilot line that will make full-size modules largely still for development purposes. We will deploy those modules in the field as well, some commercially with customers. And then as we evolve the development program, as we evolve the HVM issues in larger form factors because any time you scale from something that's, call it, 2.6 centimeters to something that is 2.5 square meters, there's issues that you're going to have to deal with through that scaling process and especially the informity of the film.

Once we have better informed around how well that's progressing, that will determine overall commercial readiness. But the entitlement here is an efficiency number that is 20-plus percent. It's an LTR that is competitive, bifaciality that's, call it, 70% and a [indiscernible] that's somewhere in the mid-teens, which is better than we have now with CdTe. That's the aspiration of what we're trying to accomplish, but a lot of work between now and then.

Your next question comes from the line of Vikram Bagri with Citi.

I wanted to ask sort of a 2-part question about India capacity and potential for cancellations. First on India, can you talk about the pricing environment? I was wondering if you can give us confidence that the sales in the market are viable, the pricing is stable and is expected to stay stable throughout the year. I understand there is no sizable spot market in the U.S., but is there a possibility some of that volume can be rerouted to the U.S. given the 15% tariffs now in place?

And then finally, on India, it seems there is a lot of domestic capacity for panels being ramped up. How do you think about that capacity in the long term and the viability of that market for first solar panels? And then as a follow-up on cancellations, you've historically mentioned pricing and EU players pulling back investments from renewables in the U.S. market has created this risk of cancellations. With lower tariffs, how are the conversations going with customers? Is that risk much lower now? And if you can quantify how much of the contracts are potentially at risk of cancellations?

All right. I'll deal with the India question first, and I don't know if you got the tariff question. Okay. All right. So look, what I would say is right now, pricing is a point to look at in India, it is lower. What I would say though we're effectively realizing high teens to low 20% gross margin. So if you want to look at it from a gross margin standpoint, I think you have to understand the cost of manufacturing in India is significantly lower than what the cost of manufacturing is in the U.S. So yes, lower pricing, but you basically balance that out with a lower cost of production and you're realizing kind of high teens to low 20% type gross margin in India.

So that's obviously where we stand right now. And if you look across the horizon of what -- how would we optimize that production. And one of the things I think you said is the risk of overcapacity. Look, I agree that there is that risk, but there's also what's counterbalancing that and why there's also robust demand for our product and technology in India is that the approved list of model and manufacturers in India is now moving further upstream. So it started off with the module itself now effectively for any project that is commissioned in April of this year or Q2, we will now have a cell requirement to be made domestically. And then there's also in -- they foreshadowed already a wafer requirement that effectively starts to come into effect in 2028.

And there's not existing vertically integrated manufacturing capability in India at this point in time. And I think what has seen -- what some of the manufacturers have seen is that it has become more challenging as they try to get into cells and then even more so if they try to get into the wafer. And then obviously, they're trying to figure out the poly side of the house as well. So a lot still needs to happen from that standpoint.

We also believe that from a cost standpoint, even if that vertically integrated supply chain were to happen, we will be cost advantaged with our vertically integrated manufacturing facility with everything being done within the 4 walls of a factory, right? So there's a cost advantage, we think. We also believe we have an energy advantage, which will be further enhanced when we implement our CuRe technology in India, which will start beginning of 2027.

So I look across the horizon, I feel like I got a better product, better technology. I've got a better policy environment that should continue to advantage us in terms of India. As a backdrop, to your point, we will continue to evaluate the construct of potentially bringing some of that product into the U.S. market, and we'll do some of that with a view of creating some competitive tension into the domestic market because I don't want customers to believe that they don't have an alternative path other than selling into the domestic market. I'd like to be able to say, sure, at the right tariff construct, we can also redirect that product and blend it in with some S7 domestic product and still command pretty good pricing.

It's a little bit choppy to continue to move back and forth because today, we primarily make a fixed tilt product for the India market and then to have to convert the factory to make a tracker product for the U.S. market or export market that's a little clunky and you incur some downtime and some cost and the transportation of bringing that product into the U.S. is more expensive than we would like. So ideally, if we can just harmonize and keep running that factory all out serving the domestic market, getting good ASPs, that's our preferred path, but we clearly would look at bringing some of that into the U.S. market if the situation and the dynamics are right.

Vikram, you asked around cancellation risk and whether tariffs are playing a part there. Look, tariffs coming down are clearly helpful, but I don't think what you've been seeing on the cancellation side over the last year has really been tariff related. So the 2 aren't necessarily linked. I mean what we said before and what we've been seeing is more of a strategic shift by certain players, especially oil and gas and the European utility players to reallocate capital away from renewable development in the U.S. into some of the more core business on oil and gas development or European utility development.

So where we've seen some of those players move away, we've seen others enter into the space in the U.S. and pick up some of that slack and see an opportunity in taking over those projects to develop them. So it's hard to handicap the cancellation risk in the backlog. Sure, it could exist. I think historically, what we've seen has been potentially more risk around the international product, just given the value of domestic content. If you look at today, the amount of international product sitting in our backlog is pretty small. You can see that by how much we're producing out of Southeast Asia this year relative to U.S. product.

And so that product has potentially been more challenged. I think the U.S. demand has been strong. And even if there were cancellations, there are much more opportunity to move any terminated U.S. product back into the U.S. market. I'd also say that clearly, we've been enforcing termination penalties and fees and making sure that we capture the value in the contract. So if there's a contractual obligation to someone to pay an amount for terminating those contracts, we've been enforcing that and we'll continue to do so.

Our final question will come from the line of Ben Kallo with Baird.

I just wanted to square time to power is a big question or big emphasis out there. And I know you have so many moving pieces. And just bookings, I'm trying to square time to power and the need for electrons with how you guys are doing bookings with your different moving pieces right now.

Look, I think Ben, I mean you got to remember, we've got a luxury -- we're kind of in a nice position. We've got 50 gigawatts of contracted volume that is going to carry us forward. There's a clear sense of urgency from customers around execution and time to power. I mean there's customers that obviously have safe harbored under Section 48 and they need to get their projects commissioned by the end of 2028. There's others that are safe harbor under 48 and those who are continuing to safe harbor because they can do that up through the middle of this year, we then get an opportunity to get projects done through the end of 2030.

So there is quite a bit of demand. Our customers are also dealing with a lot of angst of trying to figure out permitting issues and other things they're trying to do, financing issues, getting things in place for the current projects that they're executing against in construction and what have you. And then thinking through a longer position around across their portfolio, both earlier stage and then obviously, late-stage development. So that sense of urgency is there. People are being really creative looking to do more on-site generation, try to get out underneath the constraint of the interconnections.

But if a customer does have an interconnection agreement, I mean, they're running hard and they're making sure they're lining up their modules as well as their EPC agreement and balance of system equipment to make sure they can execute on time. So it's a clear catalyst. I also think it is helping us as we engage and we talk with customers around pricing as well.

There are no further questions at this time. This concludes today's call. Thank you for attending. You may now disconnect.

Good afternoon, and welcome to First Solar's Third Quarter 2025 Earnings Call. This call is being webcast live on the Investors section of First Solar's website at investor.firstsolar.com. [Operator Instructions] And please note that today's call is being recorded.

I would now like to turn the conference over to your host, Byron Jeffers, Head of Investor Relations. Please go ahead, sir.

Good afternoon, and thank you for joining us on today's earnings call. Joining me are our Chief Executive Officer, Mark Widmar; and our Chief Financial Officer, Alex Bradley. During this call, we will review our quarterly results and share our outlook for the remainder of the year. After our prepared remarks, we'll open the line for questions.

Before we begin, please note that some statements made today are forward-looking and involve risks and uncertainties that could cause actual results to differ materially from management's current expectations. We undertake no obligation to update these statements due to new information or future events. For a discussion of factors that could cause these results to differ materially, please refer to today's earnings press release and our most recent annual report on Form 10-K as supplemented by our other filings with the SEC, including our most recent quarterly report on Form 10-Q. You can find these documents on our website at investor.firstsolar.com.

With that, I'll turn it over to Mark.

All right. Good afternoon, and thank you for joining us today. Beginning on Slide 3, I will share some key highlights from Q3 2025. Since our last earnings call, we secured gross bookings of approximately 2.7 gigawatts at a base ASP of $0.309 per watt, including 0.4 gigawatts of Series 7 modules impacted by previously disclosed manufacturing issues booked at an ASP of $0.29. We terminated 6.6 gigawatts of bookings under multiyear agreements defaulted on by affiliates of BP, a European oil and gas major, at a base ASP of $0.294 per watt. As a result, total debookings since the last earnings call were approximately 6.9 gigawatts. And our current expected contracted backlog is approximately 54.5 gigawatts.

We delivered a record 5.3 gigawatts of module sales and reported Q3 earnings of $4.24 per diluted share, both near the midpoint of our previous earnings call forecast. Gross cash increased to $2 billion, supported by improved working capital, new bookings, deposits and accelerated customer payments ahead of the effective date for the new beginning of construction guidance. Alex will walk through our financial results in more detail later in the call.

From a manufacturing perspective, we produced 3.6 gigawatts of modules in the third quarter, 2.5 gigawatts from our U.S. facilities and 1.1 gigawatts from our international operations. In Q3, we reduced production in Malaysia and Vietnam, primarily due to lower demand driven by the customer default previously mentioned.

We continue to advance our domestic capacity expansion, notably at our Louisiana facility, where we initiated production runs and started plant qualification. We've also continued to pursue the enforcement of our intellectual property rights. During the quarter, we made 3 separate filings requesting that the U.S. Patent and Trademark Office or PTO, deny petitions filed by affiliates of Canadian Solar, JinkoSolar and [ Munder ] that seek to invalidate our U.S. TOPCon patents. Our filings include a reference to comments made earlier this year by the Acting Director of the PTO, who stated, "The longer a patent has been in force, the stronger and more settled the patent owner's expectations should be."

We believe our ongoing vigorous enforcement of our decade old U.S. TOPCon patents, which we consider fundamental to producing that technology, is a prime example of a patent holder having settled expectations of the integrity of its IP rights. The view that module manufacturers and their customers and financing parties should strongly consider the potential hurdles of producing, selling or purchasing modules employing TOPCon cell technology is not one held just by us. For example, earlier this quarter, the CEO of [ ES Foundry ] explained that his company's decision to focus on manufacturing PERC technology was due, at least in part, to the "legal troubles that would be encountered by TOPCon producers."

Lastly, we're pleased to continue building on our commitment to responsible solar, not simply by exceeding industry norms and sustainability and human rights, but by continuously improving on our own performance. Our Ohio facilities, which previously earned a Silver rating in the Responsible Business Alliance Validated Assessment Program, have progressed to a Gold rating in its 2025 audit, which was completed this past quarter.

Turning to Slide 4. I will now provide an update on our manufacturing operations. As it relates to our Alabama facility, 2 of our domestic glass suppliers faced manufacturing disruptions that limited our ability to operate at full capacity, which impacted Q3 production by approximately 0.2 gigawatts. The primary supply chain issue resulted from throughput limitations due to insufficient initial facility readiness at a new factory, while simultaneously, a different supplier experience unplanned downtime. Corrective actions have been implemented at both suppliers, and our U.S. glass supply base is again positioned to meet our requirements. While now resolved, this resulted in a temporary shortage of cover glass supplied to our Alabama facility, which led to reduced production and increased underutilization charges in the third quarter.

Our Louisiana factory has initiated integrated production runs, started plant qualification, and the early-stage ramp is slightly ahead of expectations. We anticipate receiving required production certificates in Q4 and will begin shipment at that time. As it relates to our international capacity, we have previously indicated the implementation of the Reconciliation Act earlier this year as well as the evolving universal and reciprocal tariff environment could potentially support a business case to establish 1 or more lines in the U.S. to finish front-end production initiated within our international fleet.

We have made the decision to establish a new production facility in the United States, allowing us to onshore the finishing of Series 6 modules initiated by the company's international factories. While the location is subject to final negotiations, we have -- with an announcement expected in the coming weeks, the planned capacity will be 3.7 gigawatts. Production will start at the end of '26 and ramp through the first half of 2027.

As we previously noted, such an investment is expected to enable additional production in the U.S. market that we expect will be fully compliant with forthcoming guidance, as well as improve the gross margin profile of our sales of reducing tariff charges and logistics costs associated with importing finished goods. Furthermore, we expect that the modules produced at this facility will provide domestic [ content points ] benefits for our customers and qualify for 45X module assembly tax credits.

We continue to evaluate options for the remainder of our international Series 6 capacity, including options related to long-term U.S. market demand, U.S. market supply and the global tariff environment. Shifting to the current policy landscape, the U.S. policy and trade environment remains generally favorable. As we have long stated, one of First Solar's key competitive differentiators is the ability to provide certainty to our customers, both in terms of pricing certainty and the certainty of timingly producing and delivering product. These attributes are particularly valuable in the U.S. solar market, where [ FIA-compliant ] suppliers who have domesticated their supply chains and localized their production capabilities provide the surest pathway to enable developers to realize tax benefits and to mitigate the exposure of project pro formas to both the imposition of tariffs and the risk of project schedules associated with relying on imported products.

A number of trade and policy developments over the quarter amplified these competitive [ initiators ]. In August, the U.S. Court of International Trade ruled that the Biden administration's 2-year suspension of circumvention related antidumping and countervailing duties was unlawful, paving the way for possible retrospective duty payments on solar imports brought into the United States between June of '22 and June of ' 4.

Also during the quarter, the U.S. International Trade Commission issued a preliminary affirmative determination in an antidumping and countervailing duty case known as Solar 4 that imports of crystalline silicon cells and modules from India, Indonesia and Laos are causing material injury to the U.S. solar industry. In addition to a range of alleged illegal subsidies, the petitioners identified dumping margins of approximately 90% for Indonesia, approximately 247% for Laos and approximately 215% for India.

Also during the quarter, U.S. Custom and Border Protection issued a notice of initiation investigation and interim measures against an affiliate of Huawei Solar in response to a claim submitted by the American Alliance for Solar Manufacturing Trade Committee, of which First Solar is a member, that Huawei had effectively transshipped Chinese solar cells and modules into the United States through India. In addition, we, together with the rest of the industry, are awaiting the results of the administration's 232 polysilicon and derivatives investigation, including the potential for incremental tariffs impacting the crystalline silicon supply chain. From a policy perspective, the industry also awaits guidance from the administration related to project impacts from foreign entity of concern, or FEOC procurement, which may be delayed as a result of the ongoing government shutdown.

In short, there continues to be mounting headwinds or uncertainties for U.S. developers associated with procurement dependent on Chinese crystalline silicon supply chain, which we believe enhances the value proposition of our vertically integrated production capabilities. It also validates our approximately $4.5 billion investment strategy of expanding our U.S. manufacturing production and reshoring supply chains, which began under the first Trump administration and continues through the current Trump administration, with our most recent facility currently ramping in Louisiana and the announcement of our new U.S. finishing line. This activity places us uniquely in the intersection of several of the administration's key priorities, including those related to domestic manufacturing job creation, American energy and energy affordability and serving among the generation solutions that enable the U.S. to win the artificial intelligence race against China.

Turning to India. Since our last earnings call, there have been several notable policy deployments. First, significantly, the application of tariff rate for imports of finished modules into the U.S. was increased to 50%. We continue to monitor dialogue between the U.S. and Indian government related to a potential bilateral trade treaty easing of tariffs between the 2 countries. As it relates to the country's domestic market, the Indian government continues to promote its domestic renewable energy value chain by progressively including cells and the remit of the approved list of models and manufacturers under a recently announced [ List 2 ]. Inclusion in List becomes mandatory for solar OEMs to sell into key segments of the domestic market, effective June of '26. Notably, First Solar was automatically qualified in this list, which was released in August of '25.

The Indian government also released a stakeholder consultation in September of '25 related to a further extension of the ALMM regulations to include domestically made wafers for potential deployment after June of 2028. Once again, First Solar's India's production is expected to automatically qualify. We anticipate that these regulations will progressively strengthen our position in the Indian market by leveling the playing field.

I'll now turn the call over to Alex to discuss shipments, bookings, Q3 financials and guidance.

Thanks, Mark. Beginning on Slide 5. As of December 31, 2024, our contracted backlog totaled 68.5 gigawatts valued at $20.5 billion or approximately $0.299 per watt. Through Q3, we recognized 11.8 gigawatts in module sales and recorded gross bookings of approximately 5.1 gigawatts. This includes 4 gigawatts booked between the enactment of the Reconciliation Bill in early July and the September 2 effective date for the new commenced construction guidance.

Since our last earnings call, we had gross bookings of 2.7 gigawatts and average selling price of $0.309 per watt. This includes approximately 0.4 gigawatts of Series 7 modules impacted by previously disclosed manufacturing issues booked at an ASP of $0.29 per watt. The remaining bookings of 2.1 gigawatts were sold into the U.S. market at a blended ASP of $0.325 per watt. As a reminder, a significant portion of our contracted backlog includes pricing adjustments that may increase the base ASP, contingent upon achieving specific milestones within our technology road map by the time of delivery, according to the ASPs presented to exclude potential adjustments related to module bin, freight overages, commodity price fluctuations, committed wattage, U.S. content volumes and tariff changes.

Our recent bookings scheduled for delivery in periods where such milestones could be met, the potential value is reflected in our backlog as an opportunity rather than the base ASP represented. For example, among recent bookings, we secured a 0.6 gigawatt order for 2027 delivery at an ASP of $0.316 per watt, with the potential for an incremental $0.046 per watt contingent on achieving specific milestones within our technology road map.

Demand in the U.S. remains strong. However, we recorded full year debookings totaling [ 8.1 gigawatts ] as of September 30, including 6.9 gigawatts in the third quarter. The majority of these were driven by contract terminations with affiliates of BP, which accounted for 6.6 gigawatts. Note, aside from the contract terminations with the BP affiliates, a number of other terminations were for project-specific reasons as opposed to reflecting customer pivots from solar project element generally. For example, our Q3 bookings include volume expected to be delivered to a customer who terminated a project in 2024, but [ is recommended to solar development ] in 2025 continues to source its module supply with First Solar. In addition, we're currently in active negotiations for the procurement of new volume with another customer who previously terminated a contract with us for a specific project of theirs earlier this year. In both cases, these customers satisfied their termination payment obligations.

In prior calls, we highlighted the emerging risk of a strategic shift concerning multinational oil and gas and power and utilities companies, particularly those based in Europe, with some moving away from renewables project development and back towards fossil fuel investments. On September 30, First Solar filed a lawsuit against BP Solar Holding LLC and its affiliate Lightsource Renewable Energy Trading following their failure to [ cure ] multiple breaches of contractual obligations. According to public reports published earlier in the year, BP has been looking to divest its interest in its renewable development arm. Despite agreements to purchase approximately $1.9 billion or 6.6 gigawatts of solar modules, these BP affiliates did not meet required payment obligations or provide required payment security. After issuing default notices providing opportunities to cure, we terminated the contract, which entitles us to approximately $385 million in termination payments. Of this amount, we've recognized $61 million in previously collected down payments as revenue.

We're seeking monetary damages which includes approximately $324 million in remaining termination payments, along with certain other receivables for solar modules previously delivered and interest. And if realized, the $324 million, will be recognized as revenue. We were already willing and able to continue fulfilling our contractual obligations to these BP affiliates and disappointed that we must resort to litigation. The modules that are subject to the contract breach are a mix of domestic and international product, most of which were scheduled to be produced in Q3 and future quarters. The delivery is expected to extend into 2029. We're working to address the planned allocation of module inventory that could have been delivered to the BP affiliates, if not for their contract breach.

With respect to such planned future module production, the market for these modules may be constrained by the U.S., Indian and European policy and market conditions discussed on the February earnings call and that has since been further exacerbated in the U.S. with our traditional utility scale customer experiencing transmission and permitting rate challenges in large part due to the constraints reflected in the July Department of Interior memo related to renewables project development, the ongoing government shutdown and the impact of tariffs. Note, these same factors, which are further exacerbated by the breach of contract to the BP affiliates, given our loss of contracted offtake for the product, may drive further underutilization charges being realized in 2026 as it relates to our Southeast Asian production facilities for the planned module volume expected to be delivered to these BP affiliates.

As a result, our quarter-end contracted backlog stood at 53.7 gigawatts valued at $16.4 billion or approximately $0.305 per watt. And as of today, our total expected contracted backlog stands at 54.5 gigawatts, excluding any volumes sold off at the end of the quarter.

Moving to Slide 6. Our total pipeline of mid- to late-stage booking opportunities remained strong, and bookings opportunities of 79.2 gigawatts and mid- to late-stage booking opportunities of 17.8 gigawatts. Our mid- to late-stage pipeline includes 4.1 gigawatts of opportunity that are contracted subject to conditions precedent. As a reminder, signed contracts in India will not be recognized as bookings until we receive full securities ahead of the offer.

I'll now cover our third quarter financial results on Slide 7. We recognized 5.3 gigawatts of module sales during the quarter, including 2.5 gigawatts from our U.S. manufacturing facilities. Our net sales totaled $1.6 billion, representing an increase of $0.5 billion compared to the prior quarter. This increase was primarily driven by higher shipment volumes and anticipated back-weighted profile of deliveries over the course of the year. Our sales included $81 million in contract termination payments, with $61 million related to the contract breach with the BP affiliates. This amount was recognized from existing cash deposits.

Gross margin for the quarter was 38%, a decrease from 46% in the prior quarter. This decrease was primarily due to a lower mix of modules sold from our U.S. manufacturing facilities, which benefit from Section 45X tax credits. Additionally, we incurred higher underutilization costs due to continued production curtailments in Southeast Asia, the BP affiliates' termination and glass supply chain disruption at our Alabama facility.

As an update on warranty-related matters, we've resolved certain obligations and advanced negotiations with additional customers regarding manufacturing issues affecting select Series 7 modules produced prior to 2025. Based on our settlement experience, the estimated number of affected modules and projections of probable remediation costs, we believe a reasonable estimate of potential future losses will range from approximately $50 million to $90 million. Within this range, we recorded a specific warranty liability of $65 million, an increase of $9 million from our prior estimate, representing our best estimate of expected future losses associated with these manufacturing issues.

As of the end of the third quarter, we maintained approximately 0.6 gigawatts to potentially impacted Series 7 inventory, including 0.2 gigawatts under contract and included in our backlog. SG&A, R&D and production start-up expense totaled $145 million in the third quarter, an increase of approximately $6 million compared to the second quarter. This increase was primarily driven by start-up costs associated with the accelerated ramp-up of our Louisiana facility aimed at providing resiliency to our U.S. production for the year. Operating income for the quarter was $466 million, which included $138 million in depreciation, amortization and accretion, $49 million in ramp and underutilization costs, $37 million in production start-up expense and $7 million in share-based compensation.

Nonoperating income resulted in a net expense of $6 million in the third quarter, representing a decrease of approximately $4 million compared to the prior quarter. This is primarily driven by higher interest income as a result of an increase in investable cash, cash equivalents and marketable securities. Tax expense for the third quarter was $4 million compared to tax expense of $10 million in the second quarter. This decrease in tax expense was primarily driven by a $19 million discrete tax benefit associated with the acceptance of the filing position on an amended tax return in a foreign jurisdiction, partially offset by higher pretax income. This resulted in third quarter earnings of $4.24 per diluted share.

Turning to Slide 8, I'll discuss select balance sheet items and summary cash flow information. At the end of Q3, our total cash, cash equivalents, restricted cash and marketable securities stood at $2 billion, an increase of approximately $0.8 billion from Q2, driven by improved working capital, new bookings deposits and accelerated customer payments ahead of the effective date for new beginning of construction guidance. As disclosed in our Form 8-K on October 20, 2025, we executed 2 Section 45X tax credit transfer agreements totaling up to $775 million in tax credits. The fixed agreement for the sale of $600 million in tax credits at a purchase price of $573 million, payable by year-end, a variable agreement with the sale of up to $175 million in tax credits, with payment expected in Q1 2026. These transactions highlight the liquidity of the 45X credit market and strengthen our near-term liquidity to support our technology road map and expansion priorities.

Accounts receivable decreased sequentially, driven by higher cash collections. Quarter end, total overdue balances were approximately $334 million, including a deferred payment settlement of $93 million with a customer, which interest payments remain current. In addition, we have approximately $70 million in uncollected receivables related to termination payments. We currently have $82 million accounts receivable for delivered modules that are aged and past due with the aforementioned BP affiliates. This does not include any additional anticipated proceeds from potential recoveries associated with their breach of contract.

Although termination payments remain contractually due, these balances are expected to persist pending the resolution of arbitration and litigation. In all instances of contract termination, we're actively pursuing all available remedies, including arbitration and litigation to enforce our contractual rights and recover amounts owed. Deferred revenue increased by $395 million, primarily due to accelerated customer payments ahead of the effective date for new beginning of construction guidance, partially offset by revenue recognized from delivered modules and termination payments. Capital expenditures totaled $204 million in Q3, mainly driven by investments in our Louisiana facility, where we initiated production runs and started plant qualification. As a result, our net cash position increased by approximately $0.9 billion to $1.5 billion.

Before addressing our updated guidance, I'd like to revisit the policy and trade environment that shapes our operational decisions throughout the year. These evolving dynamics influenced our strategy impacted quarterly performance and informed our adjustment to forward guidance. Our 2025 shipment profile acquired sustained production to fulfill contractual commitments concentrated in the second half of the year amid significant trade and tariff uncertainty. During this period, we navigated a range of potential tariff scenarios, customer negotiations and regulatory developments, including Section 232 actions, FEOC restrictions and AD/CVD investigations. At 1 point, we managed 2 possible tariff regimes, a continuation of a 10% universal tariff or adoption of reciprocal tariffs initially set at 26% for India, 24% for Malaysia and 46% for Vietnam, later amended to 50%, 19% and 20%, respectively.

Our strategy has been to maintain sufficient capacity to fulfill international module commitments and to actively pursue tariff recoveries from customers at the same time as temporarily curtailing or idling capacity and recording underutilization in circumstances where tariff recovery was unlikely and module sale economics would be challenged. The upper end of our prior guidance assumes sustained production with partial tariff recoveries, whereas the lower end reflected risk by determination related impacts, including additional underutilization costs and margin erosion from terminated contracts.

Three significant updates drive our revised guidance ranges today. Firstly, the decision announced today to establish a new 3.7 gigawatt U.S. production facility, enabling us to onshore finishing for Series 6 modules initiated by our international fleet, will result in approximately $330 million of the total program direct spend, including approximately $260 million of capital expenditures, approximately $70 million of noncapitalized expense associated with equipment de-installation, cleaning, packaging, shipping, import tariffs and reinstallation. Of this, we expect an incremental $26 million of CapEx and $2 million of production start-up expense in 2025. In addition, we forecast approximately $10 million of incremental indirect charges in 2025 associated with this decision, including severance and asset impairment expenses.

As previously noted, we continue to evaluate options for our remaining Malaysia and Vietnam facilities. Today's guidance excludes any additional costs associated with potential restructuring charges or asset impairments that may impact 2025 or future operating results. Secondly, as it relates to the termination of contracts with affiliates of BP, the loss of gross margin assumed in 2025 was largely offset by the termination payment recorded in Q3.

Increased underutilization expenses from reduced plant throughput as we curtail production given this termination of demand, were incorporated in the low end of our guidance range. Thirdly, as previously discussed, simultaneous incidents that 2 of our glass suppliers led to a shortage of glass available at our Alabama facility in Q3. This reduced full year production by approximately 0.2 gigawatts, resulting in a reduction to gross margin and Section 45X tax credits and increased underutilization costs.

Turning to Slide 9. I'll now outline the key updates to our 2025 guidance ranges, which incorporate the cascading impact of our third quarter operational and financial results. Our net sales guidance is projected at $4.95 billion to $5.2 billion, reflecting a downward revision of approximately 0.5 gigawatts from the top end of our prior guidance. This adjustment primarily reflects reduced international volumes sold due to customer terminations, partially offset by termination payments, as well as 0.5 gigawatt reduction in assumed domestic India sales following the midyear redirection of India products in the U.S. market to the domestic book and bill market, driven by the high tariff for imports into the U.S.

Additionally, U.S. manufactured volumes sold is expected to decrease 0.2 gigawatts at the high end of the guide as a result of Q3 glass supply constraints at our Alabama facility, partially offset by 0.1 gigawatts at the low end by expected increased supply from our Louisiana factory. Gross margin is expected to be between $2.1 billion and $2.2 billion or approximately 42%. This includes approximately $1.56 billion to $1.59 billion of Section 45X tax credits and $155 million to $165 million of ramp and underutilization costs. The bottom end of our previous guide has increased significantly due to further curtailment of our Southeast Asia manufacturing capacity following the contract terminations by affiliates of BP.

SG&A and R&D combined expense is expected to total $425 million to $445 million. And total operating expenses, which include $90 million of production start-up expense, expected to be between $515 million and $535 million. Operating income is expected to range between $1.56 billion and $1.68 billion, implying an operating margin of approximately 32%. This guidance includes $245 million to $255 million in combined ramp ,underutilization and production startup expense, as well as approximately $1.56 billion to $1.59 billion in Section 45X tax credits, net to the anticipated discount associated with the sale of these credits. This results in a full year 2025 earnings per diluted share guidance range of $14 to $15.

In summary, the upper end of our EPS guidance range is reduced by $1.50 per diluted share. This includes approximately $0.60 per share from the supply chain impact at our Alabama facility, which resulted in increased underutilization costs and lower volumes sold. Contract termination by BP affiliates reduces EPS by another approximately $0.60 per share due to increased underutilization costs and lower volumes sold, partially offset by termination payments. The remaining $0.30 per share is a combination of reduced India volumes sold, increased production start-up expense, finishing line costs and warranty expense, partially offset by non-BP affiliate termination payments and decreased full year tax expense. Capital expenditures for 2025 are now expected to range between $0.9 billion and $1.2 billion. Our year-end 2025 net cash balance is anticipated to be between $1.6 billion and $2.1 billion.

Turning to Slide 10, I'll now summarize the key messages from today's call. Despite some near-term headwinds, we continue to believe that our integrated domestic manufacturing platform and reshored domestic supply chain position us for long-term success. We're building a new 3.7 gigawatt capacity module finishing line in the U.S., which is expected to begin production in Q4 of 2026, and ramp at the first half 2027. We delivered a record 5.3 gigawatts of module sales, and our Q3 earnings per diluted share came in above the midpoint of our guidance range at $4.24 per share.

We saw an improvement in our gross cash position to $2 billion and recently executed agreements to sell additional Section 45X tax credits, which we expect to further enhance our liquidity position. We revised our full year guidance to reflect the impact of third-party glass supply chain disruptions as well as the termination of 6.6 gigawatts of volume by affiliates of BP, which we recognized a partial termination payment and have filed a lawsuit for damages for breach of contracts.

With this, we conclude our compared remarks and open the floor for questions. Operator?

[Operator Instructions] The first question comes from Philip Shen, ROTH Capital Partners.

First one is on the 6.6 gigawatts of termination with BP. Just want to check in on whether or not in terms of rebooking this volume, it sounds like it's volume from '26 through '29. What kind of incremental pricing do you think you can get for this? Would you expect these bookings to get locked in following the Section 232 tariff announcement, which should be near term? So sometime in Q4 or Q1? And then -- or do you think you might wait until things fell down post 232?

And then the second question is tied into this, as it relates to the 232. Is there room for negotiation, you think, with any of your fixed price contracts that you have out there, where they may not have been accounted for in terms of this new tariff? So just curious if you can share some color on that as well.

Yes, Phil, look, I mean, now with the termination, we clearly are going to be engaging, looking -- given our overall pipeline of opportunities, to figure out the right opportunities for the volume in the respective windows that was anticipated to be delivered. We will continue to be very patient in that regard, assuming we can get good prices. Like if you look at the 1 deal that Alex included in his prepared remarks, the base price plus the [ curators ] get that number into a little bit north of 36, close to $0.365, and I think that's a number that we would continue to look to engage. But at this point in time, I think there's other catalysts that could put a little bit more momentum behind that pricing as well, especially with the 232, as you referenced, and there's still obviously, FEOC guidance that's going to continue to be provided as well.

So a lot of insights or information that still is valuable to us to gain. If we can get good pricing, we'll continue to layer on some volumes into the years that we currently have available supply. But I think the value of being patient here is going to only work to our benefit in that regard.

As it relates to the fixed-price contracts, the value of certainty, I think, is what Alex indicated in his comments and we said that many times before. The contracts are -- do not have latitude for something like a revised tariff environment that was not assumed at the time of the committed obligations that both parties assumed. So they do not allow [ openers ] for 232s, as an example. But we still have capacity in the foreseeable future, especially through our international operations that we can use to engage the market and provide supply once we know the outcome of 232. But yes, the existing contracts that are on the books right now, those are obligations for both parties, and we take that seriously.

That's also why we took the position that we did with the Lightsource BP transaction and the termination enforcing our contractual rights. We worked, as we indicated in our prepared remarks, to try to get to an outcome that would be beneficial for both parties. We couldn't get there. So we had to enforce the contract. And we hold ourselves accountable to that as well. We have contracts and obligations to deliver. Pricing is fixed, except for certain respective adders and would not include tariff-related outcome or any other adjustments that were a result of the 232s that are being currently under investigation.

The next question today is Brian Lee from Goldman Sachs.

I guess, first, I just want to make sure I interpret this correctly. It sounded like, Mark, you're saying given the adders indicative pricing, $0.36, $0.365 per watt, that's maybe kind of level of entitlement you think you'll ultimately settle that once this game of patience evolves to when you really engage in pricing discussions post FEOC and 232? And then the second question, just on -- the 3.7 gigawatt finishing line, great to hear on that. But is the CapEx all being spent this year? And then maybe high-level thoughts around just expanding that? Why not simply do a full 7 gigawatts plus to come both to Vietnam and Malaysia volume capacity?

Yes. So Brian, I think as you summarize what I said to Phil, I think that's the objective of where we'd like to ultimately see, especially with the -- on the other side of understanding of FEOC and the 232. That's kind of the entitlement that we would expect with -- especially for the new technology and the value add that we provide through CuRe. So I think you summarized that well. I'll let Alex talk to the CapEx.

But before that, as it relates to where we are right now is 3.7. One of the things that we do want to try to keep measured is -- the finishing line will bring with it domestic content, right? But it's not going to bring the entire value stack of domestic content that we capture through our production in Perrysburg. The front-end semi-finished product that comes into the U.S., obviously, by definition, will not value -- not create domestic content value. So what we're trying to do is keep that thought pretty much balanced so we can continue to blend.

So even that contract that I referenced with the adders that got into the mid-36, that was still a blend of international and domestic. And so we think that by keeping that balance, that allows us to realize the highest potential value for that finishing line. So that's where our head is right now. 3-7 kind of balances very well with the production that we have in Perrysburg, which is north of 3 gigawatts as well. We'll continue to evaluate whether there's an opportunity to bring more into the U.S. using the front-end capacity we have internationally. We'll have opportunities to better reassess that once we understand the outcome of 232 in particular and the FEOC guidance that we're looking forward to, and we'll make that decision at that time.

And Brian, just on the spend. So what we said is about $330 million of direct spend. Of that, $260 million is CapEx. And of that $260 million, we'll spend about 10% of it this year. So $26 million, the remainder will be spent in 2026. The other $70 million, so $260 million of CapEx, $330 million of total spend. The other $70 million is noncapitalizable spend. So that's going to be decommissioning of the current tools, taking them out, cleaning, packing them, the freight to get them to the U.S., some tariff on the import, reinstallation. So all that will be expensed versus capitalized. Of that $70 million, we're only forecasting spending about $2 million this year, the rest will come in 2026.

There is some incremental charge that we'll hit this year. We said about $10 million. That's indirect associated with what we're doing. So it's not part of $330 million. That's some severance for some associates that will be impacted in Southeast Asia. And then there'll be some equipment write-off as well. There may be more associated with that in 2026, and we'll give you more color on that when we guide for next year.

Your next question comes from Moses Sutton from BNP Paribas.

In the past, Alex, you delineated, I think, 85% of either gigawatts or customers were in like a true take-or-pay structure contractually. And 15%, maybe it was 16% were supported by the nonrefundable deposits or termination fees. Was BP in the latter bucket? Hence, that you're going after and litigating for that? Given BP is over 10% of the backlog or was at least, I would assume that they weren't in the take-or-pay bucket, but I just want to confirm and if you can comment on which bucket they are? And can you update how firm the rest of the contracts are? I think it would be a good time to give a mark-to-market on that.

Yes. So when you say take or pay, I think maybe what you're referring to is termination for convenience, potentially. And so correct me if I'm wrong, but if you're referring to that piece, then the BP contracts were not contracts that had an ability to terminate for convenience. So they had no ability to exit those contracts. Now if they had wanted to cancel, they could have certainly worked with us, we would have had a discussion and potentially a solution we could have come to. But as Mark said, unfortunately, despite working with them for a long period of time, they chose to default on these contracts.

We did have some cash deposits from them. That's the piece that we recognize as revenue associated with the termination. We also had some LCs. Generally, that was going in some of the accounts receivable we had outstanding, so we have pulled those LCs as well. And then the residual is generally parent guarantees, and that's the piece that we will be litigating to recover.

The next question comes from Jon Windham, UBS.

Just a quick point of clarification, and then I'll get on to the [ fibro ] question. Was the cancellation related to BP, was that all from international factories?

No. It was a mix of products, both international and domestic.

Just clear on this and to -- the current year supply was essentially all international. So it was a mix. Again, the contract goes up multiple years with deliveries anticipated to go out through '29. So think of it as the front of that is mostly international. And as you get more longer dated, it then transitions into domestic.

And then so just thinking about sort of net net, is it half half, how should we think about it?

Yes. I mean it's more than half of it being domestic, but it's a very -- a significant chunk of it being international.

Up next, we'll hear from Julien Dumoulin-Smith from Jefferies.

Just following up a little bit on the earlier commentary about the CapEx. You suggested there may be 1 or more lines. Can you elaborate under what conditions you would look to seek to open multiple new lines on the finishing front? And how you would think about that in terms of the sourcing front as well? International?

Yes. It's also a distinction of how to refine it. So right now, there will be 2 lines in -- that we'll be bringing into the U.S. for finishing. So there'll be 2 finishing lines. Okay? That is 3.7 gigawatts of capacity. We could bring more lines in, right? It doesn't have to be another 3.7. It could be effectively half of that to be another line. Or we could potentially bring in 2 lines if need be. It's something that we're continuing to evaluate. There's enough front-end capacity to enable more finishing here in the U.S.

Obviously, number of variables, number of items that we've already referenced will inform our decisions around that. We're very excited about getting the first 2 lines, which adds up to the 3.7 capacity up and running here as we exit next year. And as we continue to evaluate market opportunities and demand, then we will form our decisions, do we make additional investments and how do we bring those lines in, in terms of timing. And do we do just only 6 or do we also look potentially to bring in Series 7 as well.

Next up is Ben Kallo from Baird.

Thank you. Just following up, I think, on Brian's question earlier on pricing, the 4.1 gigawatts of opportunities confirmed but not booked. Can you talk anything about pricing there? And then with your cash balance, how do you think about that, maybe just the priorities of [ task ] going forward over the next 2 years? I know there's a lot of uncertainty.

Yes. On that 4.1, Ben, the -- that's more historical, I would say, pricing. Some of that's India. That's contract that we don't count or the booking until we receive all the security. And some of that is kind of variable pricing dynamics that we have with customers. Effectively, they can flex up or down from their MSA, their module sale agreement. So I wouldn't say that, that's really a reflection of kind of current market pricing. All I would say is that we're happy with the market pricing that we're seeing. We believe there could be additional tailwinds that could further support a very favorable pricing environment for us and continuing to gauge the market and react accordingly.

Yes, Ben, as it relates to cash, clearly, cash positions increased quarter-over-quarter. We saw some activity during that safe harbor window where we saw some volume that was 100% prepaid. Some of that was taken at the same time within the quarter, some not. So you saw the deferred revenue amount increase. We also had some improvement in the working capital position, which we talked about expecting to improve as we got further into the year.

So an increase in cash, no doubt. We're announcing some more CapEx for next year. As Mark said, we'll continue to look at additional finishing lines if there's opportunity there. But the overall framework we use to evaluate cash is one we've talked about before, it hasn't fundamentally changed around running a business day to day, looking at additional capacity, looking at M&A, especially as it relates to R&D. And then if we get to a point where we can accretively deploy that capital, we look at capital return, we'll give a further update as we go into next year's guidance, how we think about capital structure longer term.

David Arcaro from Morgan Stanley has the next question.

I was just wondering if you could give a little color on your confidence level in the 54.5 gigawatt backlog now? Are there other customers that I think could be at risk that you're aware of that you're risk weighting in there or any other market dynamics that make you think or customer-specific dynamics that make you think this debookings pace could continue or not?

Yes. So -- we've been saying now for, I don't know, it could be going on close to 2 years now, something along those lines. There's been indications by a number of large oil and gas, multinationals, international companies that are continuing to evaluate their commitment to renewables, right? And obviously, BP falls in that bucket. There's been others as well. Just think about nat grid, nat grid, obviously, a large European company that made a decision to sell down its development business, going back to now, [ Jeronimo ], sold it over to Brookfield. You could look at now as another example of a commitment to the U.S. market that has been reevaluated now. I think they've changed their perspective in that regard.

And there's a couple of others, which I won't name, but EDF is, I guess, maybe another one I would throw into that bucket a little bit. I mean, it's not oil and gas, but obviously, a large European company that's reevaluating its commitment to the U.S. market. So there's -- obviously, that risk profile is something we foreshadowed. It's something that has played it self out. If you go back and if you look at what's in our contracted backlog, you can go back and look at announced deals that we've done, who some of our larger partners are. You're going to find that, that profile is dramatically different with what sits in our contracted backlog.

Now having said that, I mean, we all know that a number of developers and IPPs here in the U.S. I mean, they're working through a number of challenges, right, and permitting issues and project-related issues and what have you that things could evolve in such a way that at a project level, we could potentially see some movement. We said in the call today, we had a couple of customers that have project-specific terminations. One of them who terminated last year, project specific, and then now they're back in our order book for more than 0.5 gigawatt of volume. And then we had another one who terminated this year that we're actively negotiating a meaningful contract with.

So I don't want to give an indication of there may not be further terminations. But I also want to somewhat reflect that I don't think something as large and structural as what we saw with Lightsource is a high risk, but at any point in time, things can evolve, things could change. A number of our partners have [ sponsor capital by ]. If Brookfield decides to go a different direction, if KKR decides to go a different direction, if TPG decides to go, Macquarie, I mean, you name whoever sponsor you want to say is behind the portfolio business. If they decide to pivot and go a different direction, I mean, there's always an inherent risk in that regard.

But what I would say is that while there's still challenges and issues that are being dealt with, there's an opportunity here. The policy environment, I think, is very -- still very positive, what came out of the One Big Beautiful Bill. There is a need for more electrons on the grid. The load profile is only going to continue to grow. And project economics and PPAs are still strong, right? So I think those fundamentals, I think, still, I would say it's enduring, and that we would have a higher level of confidence in contracted offtake agreements that we have on our books right now. But I also want to be balanced in understanding that there could be some amount of risk. But I do think that on balance, there's a lot of market opportunity for our partners and obviously, for us to continue to supply into the market.

Next up is a follow-up from Jon Windham, UBS.

Perfect. Thank you. I want to ask what the topic we haven't covered much in this call is how the [ brand ] and product quality is in Louisiana and Alabama. Can you just touch on how that's running next to expectations?

Look, the ramp for DRT, I would say that has gone well. It's an aggressive ramp that we've had -- sorry, Alabama and [ referred to it front ]. It's actually has gone well, but it's also had its own set of challenges that we've been working through in terms of the ramp process and getting to the full entitlement and throughput. And where I see the factory app right now for Alabama, I see it at a very good level. It's hitting its throughput requirements. It struggled, as we indicated in our prepared remarks with disruption on our glass supply chain. And obviously, that had an adverse impact on the factory.

Louisiana is going extremely well. We're in the midst of going through our product qualification. And that will be complete here in Q4, and we'll start shipping product. And at [ ramp ] right now, the ramp is ahead of schedule, which is all very positive for us in that regard. As you said, I think you may have mentioned product quality and the like. We are continuing to do, as always, being very diligent as we manufacture our product and to ensure that we have a high level of indication of field performance based off of not only accelerated live testing, but obviously sales deployment as well. And it's something that our level of rigor and intensity around that is only going to continue to be more heightened as the result of the initial launch of Series 7. Again, that was the launch of a new product. In this case, both Alabama and Louisiana are replications of the factories that we launched our Series 7 technology from. And the key learnings that we captured from that launch and some of the changes that we've already communicated that we needed to make to our manufacturing process, both were implemented into Alabama and Louisiana before we started production. But it's something we know with the reputation is a brand issue, we got to stay on top of it, and we're going to continue to do everything we can to meet our customers' expectations in that regard.

The next question comes from Vikram Bagri from Citi.

Just a quick question. Mark, can you remind us if there is a precedent of successful litigation against a customer who is in a similar breach of contract? Or just this case with BP will set a precedent that -- for future?

Yes. I don't have my GC in the room right now because I could ask them that question. But what I can tell you is that we are using outside counsel. We have -- we believe very strong contracts that enforce the rights and obligations of both parties. And we believe that either of those parties are in default and there's consequences associated with that. I would also use -- whether there's legal precedence, and I'm sure there are. While I can't cite them to you right now, what I would go back to is if you look at the -- I believe we've had a number across the last couple of years, somewhere in the range of north of $200 million, $250 million or so of various terminations. I think we've also disclosed that about $70 million is sitting outstanding.

Okay? That means that the vast majority of that -- those terminations were paid. Because the counterparties understood the obligations and terms of conditions of the agreements, which are essentially identical across our contracts. And they have honored that obligation in respect of that obligation and they've remitted payment. They would not have done that unless they thought -- if they thought that there was a reason why, underneath the contract, that they would not have an obligation to First Solar for their default.

So I can use 2 data points. One is just look at experience. And the other is the input that we're getting from outside counsel around our contracts. And we feel very good about the contracts, the way they're structured, the foreseeability of the contracts. And my understanding is that, again, the -- this will be the filing of this litigation is in the State of New York. I think my understanding is the State of New York has taken a very strong position around this type of condition underneath the contract for default and associated with termination payment. And generally, this -- the courts of New York have cited with the plaintiff and the situation of similar circumstances. So that's about as much information as I have. I do believe though, we're in a strong position.

Our final question today will come from Joseph Osha, Guggenheim Partners.

Hello. As we think about the timing of the finishing fab coming up in the U.S. and what the commercial environment looks like, I'm wondering what conclusion we can draw about underabsorption of Malaysia and Vietnam next year? And perhaps to put a sharper point on that, is there any market at all for products being shipped directly out of either of those 2 fabs?

Yes. So 1 thing to remember is that we're using the front-end capacity of our international facilities in order to fund that into the U.S., right? And when you think about the cost structure and absorption, especially around the capital intensity of the equipment, it largely sits on the front end of the processing. So you're going to see reasonably good absorption for that front-end manufacturing that then is finished in the U.S. We also identified that we have taken some headcount reductions. So we are minimizing the back end processing of -- and labor associated with that. And then those tools that are being used in the back end are being brought into the U.S. So therefore, their depreciation there will be absorbed against the finishing processes that are being done here in the U.S. So just to put that in perspective.

Yes, as it relates to the balance of that production, 1 of the things we're continuing to work through and we are in negotiations with a couple of counterparties to almost do a bilateral for that offtake of that volume and to structure a deal around that, assuming we can get to terms. We'd like to find potentially, a couple of large customers with large offtake requirements that we can then sort of just sole source that into those opportunities. But clearly, we believe there is an opportunity, subject to the tariff environment, subject to what happens with 232, subject to FEOC guidance and everything else. So there's some more triggering events that would have to happen.

I think we said in our prepared remarks, we have something like 6 gigawatts of contracted backlog or something like that for Series 6 international, still. So we've got some runway in terms of volume and absorption for those production assets, and then we'll continue to evaluate them as we learn more about some of these policy decisions that will be made.

Everyone, that does conclude our question-and-answer session. This also concludes our conference for today. We would like to thank you all for your participation today. You may now disconnect.